ArticlePDF Available

Convolutional neural networks for archaeological site detection -Finding "princely" tombs



Content may be subject to copyright.
Convolutional Neural Networks for Archaeological Site
Detection Finding “Princely” Tombs
Gino Caspari*
Department of Archaeology, University of Sydney
Institute of Archaeological Sciences, University of Bern
Pablo Crespo
Ph.D. Program in Economics, The Graduate Center, City University of New York
Creating a quantitative overview over the early Iron Age heritage of the Eurasian steppes is
a difficult task due to the vastness of the ecological zone and the often problematic access. Re-
mote sensing based detection on open-source high-resolution satellite data in combination with
convolutional neural networks (CNN) provide a potential solution to this problem. We create
a CNN trained to detect early Iron Age burial mounds in freely available optical satellite data.
The CNN provides a superior method for archaeological site detection based on the comparison
to other detection algorithms trained on the same dataset. Throughout all comparison metrics
(precision, recall, and score) the CNN performs best.
Keywords: CNN; object detection; archaeological remote sensing; convolutional neural net-
works; site detection
*Email: .
The archaeology of the Early Iron Age in the Eurasian steppe deals with a vast and archae-2
ologically unexplored space between Eastern Europe and Mongolia. Despite the amount3
of research which has been conducted by scholars of the former USSR and the recent wave4
of new research coming out of these areas, a quantifiable understanding of the wealth of5
cultural heritage the Eurasian steppe harbors, has yet to be achieved. One of the problems6
which hinders researchers in gaining a wider understanding is the fact that the ancient7
cultural phenomena of the Early Iron Age did not neatly adhere to modern nation state8
borders (Figure 1). The current administrative, linguistic, and institutional fragmentation9
of this vast ecological zone –the steppe–makes research on the ground difficult. Remote10
sensing in combination with automatic or semi-automatic approaches for object detection11
have been established as a tool which largely disregards these problems and is able to pro-12
vide the basis for solutions (Caspari et al.,2014). Rooted in archaeological field research13
we combine open source data with convolutional neural networks (CNNs) in order to14
encompass the newest technological advances and use them to detect elite tombs of the15
Early Iron Age in the Eurasian steppe.16
When it comes to restrictive access for foreign researchers, the Xinjiang Uyghur Au-17
tonomous Region is maybe the most extreme example in the region. It is known for its18
political and ethnical issues (Clarke,2008) and recently received international media at-19
tention due to its increasingly oppressive counter-terrorism campaigns. (Roberts,2018)20
Notoriously hard to receive permits for archaeological fieldwork in the first place, spo-21
radic eruptions of ethnic conflicts between the Uyghur minority and Han Chinese major-22
ity in southern Xinjiang can abort long-planned projects last minute. Militarized border23
zones geographically curtail the areas archaeologists can work in. Even receiving a permit24
is not necessarily a guarantee that a field campaign can be conducted as planned, since25
the security apparatus is suspicious of any research activity by foreigners. Remote sensing26
mitigates these problems of access and the quality of publicly available high-resolution27
satellite data for Xinjiang has increased dramatically over the past years (Caspari,2018).28
Figure 1: The area of interest in the eastern Central Asian steppes
CNNs have become the standard tool in computer vision applications in recent years.29
Their particular use in pattern and shape recognition is noted and popularized with the30
LeNet-5 architecture for recognizing handwritten digits (Lecun et al.,1999). Their particu-31
lar usefulness is predicated on their ability to take inputs in the shape of multidimensional32
matrices (tensors), allowing them to work with patterns in multiple directions. Pixels ad-33
jacent to each other have influence on what is identified. Most other machine learning34
algorithms used in image recognition work with inputs that take the shape of single row35
vectors, eliminating the ability to harness the information given by adjacent pixels in36
an image that are not in the same row (the pixels right below, above or set diagonally).37
Hence, CNNs are much more sensitive to identifying subtle patterns in images.38
CNNs are a versatile solution to a plethora of problems in archaeology which works39
well when plenty of data is available. It comes at the cost of not being able to fully and an-40
alytically understand the process of solving the problem. The outcomes however can be41
qualitatively assessed and the solution is reproducible. Consistent with their versatility,42
CNNs have been used in different archaeological sub fields and for a diverse number of43
tasks from sex determination of skeletal remains to solving mapping tasks and extracting44
pottery depictions from archaeological publications.45
Unsurprisingly, being one of the main categories of archaeological material, research46
on ceramics has seen a wide application of CNNs already. From recognizing vessels47
to classifying ceramic form, to understanding and classifying the structure of ceramics,48
CNNs have been useful in solving complex problems. (Benhabiles and Tabia,2016) build49
a CNN to design local descriptors for content-based retrieval of three-dimensional (3-D)50
vessel replicas. (Pasquet et al.,2017) use a CNN to detect amphorae in an underwater set-51
ting, correctly mapping around 90% of the vessels. (Hein et al.,2018) automatically extract52
and classify ceramics based on textures. (Chetouani et al.,2018) enlist the help of a CNN in53
order to classify shards and understand the movement of potters. The ArchAIDE project54
experiments with CNNs to create an as-automated-as-possible tool for the classifications55
and interpretation of shards (Gualandi et al.,2016). A similar application is envisioned56
by (Tyukin et al.,2018) with the project Arch-I-Scan which aims to automatically classify57
Roman pottery.58
Interpreting other archaeological classes of information with CNNs is still in its in-59
fancy, but a number of examples can give the reader an idea of what might be possible60
if expertly human labeled datasets are combined with CNNs. (Byeon et al.,2019) au-61
tomatically identify and classify cut marks on bones. The authors manage to demon-62
strate that CNNs recognize and classify marks with a much higher accuracy rate than63
human experts. CNNs also perform exceptionally well when tasked with determining64
the sex of skeletal remains based on CT scans thereby eliminating human bias (Bewes65
et al.,2019). In the analysis and interpretation of ancient scripts, CNNs are also begin-66
ning to make an impact. First attempts have been made in indexing Mayan hieroglyphs67
(Roman-Rangel and Marchand-Maillet,2016;Can et al.,2018) and creating a standard-68
ized corpus of graphemes for the Indus Valley script (Palaniappan and Adhikari,2017).69
Further applications of CNNs in classifying, transcribing, and ultimately translating e.g.70
cuneiform are to be expected.71
CNNs have so far found the widest application in the area of archaeological remote72
sensing. This subfield of archaeology has the advantage of already working within a73
data-focused framework where classification and mapping tasks are common. The ap-74
plication of CNNs thus comes as an obvious extension of existing automated and semi-75
automated methods. Especially with LiDAR data collection, the data volume is becoming76
too large to be analyzed through a manual approach. CNNs help to mitigate this problem77
while simultaneously maintaining a consistent approach. (Trier et al.,2019) present a case78
study mapping a number of archaeological object classes on an island in Scotland based79
on airborne laser scanning data. (Guyot et al.,2018) detect Neolithic burial mounds in80
a LiDAR-derived digital elevation model. (Kramer et al.,2017) combine aerial imagery81
and LiDAR data to detect archaeological structures using previously identified archaeo-82
logical sites as training data. Other non-invasive methods like geophysical prospection,83
in particular ground penetrating radar (Travassos et al.,2018;Ishitsuka et al.,2018;Pham84
and Lef`
evre,2018), have also seen the application of CNNs. Our own case study in this85
paper belongs to the wide field of CNN applications which arose from image processing86
conceptually close to well-known and widely applied tasks like the recognition of faces87
and vehicles in images. CNNs can be useful in any area where remote sensing data needs88
to be searched for archaeological structures. The efficient processing of image data even89
allows for real-time decision making so that (Rutledge et al.,2018) are able to present an90
autonomous underwater robot system, which allows for the autonomous surveying of91
underwater sites including path planning and acquisition of high-resolution sonar data.92
Even art historical classifications and comparisons are supported by CNNs. With the93
appropriate amount of data, it becomes feasible to define stylistic affinity. First applica-94
tions can be seen in the classification of wall paintings in Pompeii (Schoelz,2018) and95
(Li et al.,2018) approach towards dating the Mogao Grottoes wall paintings based on96
drawing styles defined by a CNN. (Wang et al.,2017) use CNNs for defining similari-97
ties of Bodhisattva head images at the Dazu Rock Carving site and thus contribute to98
the reconstruction of some of the damaged rock carvings. An application of CNNs in the99
restoration of damaged archaeology can also be seen in a paper by (Hermoza and Sipiran,100
2017) where the authors try to predict the missing geometry of damaged archaeological101
objects opening a promising avenue of research into computer-supported reconstruction102
and restoration of archaeological artifacts.103
Wherever the exploration and analysis of large data sets is aided by recognizing com-104
plex patterns, CNNs can be helpfully employed. This leads to creative applications like105
a study by (Graham,2018). The authors identify sales of human remains on social media106
platforms using CNNs to detect patterns allowing for the classification of a combination107
of images and text ultimately aiding the reconstruction of sales networks.108
The Dzungaria Landscape Project, first established in 2014 (Caspari et al.,2017), relied on110
a large-scale automated survey by means of a trained Hough Forest algorithm (Caspari111
et al.,2014). Since then, machine learning has made enormous progress and the quality112
of the freely available satellite imagery has increased substantially. Through an intensive113
on-ground survey, the project was able to obtain a dataset of archaeological structures in114
the foothills of the Chinese Altai Mountains. Accumulations of very large Early Iron Age115
burial mounds early on caught the attention of the researchers (Figure 2and Figure 3).116
It soon became clear that the southern Altai Mountains, in particular the area around117
Heiliutan were a focus of intense funerary building activity, especially during the first118
millennium BCE (Caspari et al.,2017). A number of different Early Iron Age material119
cultures in the first millennium BCE can be identified (van Geel et al.,2004). Here, we are120
specifically focusing on the funerary architecture of the Saka culture due to its relative121
homogeneity. There is a plethora of architectural remains from the Early Iron Age present122
in the survey area, but many of them are too small to be reliably detected in open source123
optical satellite data (Caspari,2018). By far the most dominant anthropogenic features of124
the landscape are large burial mounds with circular ditches around them.125
These monuments of which 59 (Caspari et al.,2017;Caspari,Forthcoming) were mapped126
Figure 2: Map generated during the 2015 survey of the Heiliutan Valley in northern Xinjiang. Large Saka
burial mounds tend to cluster. Dark grey = mound. Light grey = ditch.
during the field surveys, bear a striking resemblance to so-called Saka burials from the127
Semirechye (eastern Kazakhstan), the northern Tianshan and the Ili Valley. The term128
“Saka” is a relatively unspecific ethnic term stemming from Persian sources as (P’iankov,129
1994) elaborates and thus should only be used with the appropriate care. Over decades of130
archaeological research in what is now eastern Kazakhstan, the term, however, has come131
to denote a specific Early Iron Age material culture and is seen as a technical term among132
many researchers without implying the potentially problematic ethnic connotations. The133
Saka material culture in eastern Kazakhstan is dated to the 7th/6th cent. BCE and the134
3rd cent. BCE (Parzinger,2011). Saka burials have so far mainly been known from the135
Semirechye (Davis-Kimball,1991;Gass,2011;Nagler,2009;Nagler et al.,2010) and have136
only recently been compiled in a large study by (Gass,2016).137
The connections of Saka-related material culture into northern Xinjiang have been ana-138
lyzed (Davis-Kimball,1991;Chen and Hiebert,1995) but due to the fragmentary nature of139
archaeological data in Xinjiang have been assumed to mainly be confined to the western-140
most stretches of Xinjiang, namely the Ili Valley and the northern Tianshan. Older Chinese141
research has looked at these connections from the eastern side (Wang,1985) working on142
a number of sites which show clear relations to eastern Kazakhstan like Tiemulike (Insti-143
tute of Archaeology of the Xinjiang Academy of Social Sciences,1988), Dacaotan (Institute144
of Archaeology of the Xinjiang Academy of Social Sciences,1985), and Zhongyangchang145
(Institute of Archaeology of the Xinjiang Academy of Social Sciences,1986). The architec-146
tural features of the mounds in the Heiliutan Valley, however, suggest a strong cultural147
connection during the middle of the first millennium BCE all the way into the foothills of148
the Chinese Altai Mountains.149
Figure 3: Architectural features of Saka burial mounds.
The large burial mounds of the Saka material culture usually were built from a mixture150
of pebbles, larger round stones and earth from the alluvial terraces. Mounds are typically151
elevated and surrounded by circular rings of stones or circular ditches (Figure 3). Both152
ditch and mound are clearly visible in open source optical satellite data. The profile of153
the Saka burial mounds typically shows steep sides (sometimes three steep sides and one154
with a gentler slope) and a flat top. Maximum diameters in the Heiliutan area are typi-155
cally between 15.5m and 34.1m (89.5 %) and therefore well within the range of detectable156
objects in open-source satellite imagery (Figure 4). A group of outliers has diameters of157
over 40m. The average diameter of Saka mound in the Heiliutan Valley is 27.93m (median158
The heights of these burial mounds average at 1.97m (median 1.4m). The largest160
Figure 4: Scatterplot of Saka burial mound diameters, notice the cluster of extraordinarily large mounds
which clearly set themselves apart from the smaller ones. These “princely” tombs are easily recognizable
in open source remote sensing data.
mounds have a height of up to 6.5m. Both diameters and heights of Saka burial mounds161
in the Chinese Altai are comparable to Saka burial mounds from Issyk, Kegen and other162
cemeteries with princely tombs (Gass,2011;Samashev,2007). The Saka burials of the163
Heiliutan area are all practically identical in their composition of building materials and164
the profile of the mound. The largest mound has a diameter of 53.5m, a height of 6.0m,165
and the circular ditch measures 91.5m across. This type of burial usually has a 5:3 ratio166
between circular ditch diameter and mound diameter which again matches Saka buri-167
als from the Semirechye (Gass,2011). The large accumulation of Saka burials (Figure 2)168
with a length of almost 2km are visible from afar and one of the dominant archaeological169
places within the landscape of the Heiliutan Valley. One of these monuments has been170
excavated in 2016 by the Institute of Archaeology of the Xinjiang Academy of Social Sci-171
ences but has yet to be published like many other burial mounds in the area of interest172
the grave was unfortunately looted.173
CNNs are a specific type of neural network architectures popularized by (Lecun et al.,175
1999), which can take grid-like inputs. Our particular case is a two-dimensional grid of176
pixels, in which each pixel can be considered a source of information in the same way as177
a cell in a row of tabular data would be. Note that images can be interpreted as numerical178
grids if each pixel on each channel (RGB) is given a numerical value based on the intensity179
of the color from 0 to 255. In order to understand how CNNs work, we will define them180
as the junction of three different operating components as types of “layers”:181
convolutional layers182
pooling layers183
fully connected layers184
The convolutional and pooling layers are used to identify and summarize patterns185
in the data. The fully connected layers are used to utilize these summaries as inputs186
of a classification problem, helping us make the determination of whether our (in this187
case) image belongs to a specific class based on the model. An example diagram of these188
architectures is presented in Figure 5.189
Figure 5: CNN architecture example diagram
The network uses convolutional layers to detect simple features or patterns in the data.191
The patterns can be small and simple, but the combination of multiple simple patterns192
allow for the search of complex forms.193
Each convolutional layer is composed of two stages: convolution and detection. In the
first stage a set of convolution operations are run on the input grid. A kernel or filter is
moved sequentially on the input generating outputs on each position they take. These
are defined by:
where hi,jis the output of the convolution at position (i,j),xi+k1,j+l1portion of the input194
grid over which the filter is applied, wk,lis the filter at position (k,l), and mdetermines195
the height and width of the filter.196
Hence, the filter is a weighting square grid, which is applied to the larger input grid to197
highlight specific patterns within it. The higher the value of a convolution operation, the198
higher the chance that the pattern that the filter searches for is found. Figure 6highlights199
this process by exemplifying it. We can then use different filters to find different patterns.200
For example, using a filter of the form:201
0 1 0
0 1 0
0 1 0
would be used to identify vertical lines.202
Filters can be, and often are, initialized at random to pick on many and varied subtle203
patterns within the input grid. Each convolutional layer runs several filters on the inputs204
and outputs grids for each.205
For the detection or activation stage, the results from the convolution stage are taken206
Figure 6: Filter applied over a matrix
and passed through a function. We used the ReLU (Rectifying linear unit), which is de-207
fined as:208
σ(x)=max (0, x)
This specific function grants extra weight to all of the non-negative units. Since the209
filters can have negative values, this activation allows for extra salience of patterns.210
After activation, the outputs of the convolutional layer are used as inputs for the pool-211
ing layers.212
3. 2 POO L IN G LAYE RS213
A pooling layer summarizes the resulting activated grids through max pooling. This214
work uses max pooling. A new grid is constructed from each activated grid by assigning215
each entry of it to the maximum value of 2 ×2 subgrids. An example is shown in Figure 7.216
Figure 7: 2 ×2 max pooling
At this point, the practitioner has two choices: to summarize the results once more217
through a fully connected layer (see Section 3.3) or to repeat the process of passing the218
outputs through a convolutional layer and pooling layer once again. This is what is meant219
by making a network ”deeper.” Passing the data through extra convolutional and pooling220
layers allows for further and more subtle evaluation of patterns. This is said to elevate221
the complexity of the model.222
Fully connected layers have the basic structure of artificial neural networks or multilayer224
perceptrons. Their task is to take the outputs from the last pooling layers and classifying225
them into specific categories. Before passing the grids resulting from the pooling layers226
to the fully connected layers, the grids are ”flattened.” Meaning the results from all the227
resulting grids are combined into a single row vector. The resulting elements of the vector228
produced after the flattening are then linearly combined. This means they are written as:229
where β0is called the ”bias” and the rest of the βiare called the ”weights.” Each of230
these linear combinations is passed through an activation function yet again generating a231
single number output. This particular structure of operations constitutes what is called a232
“neuron.” The set of these activated linear combinations is called a ”hidden layer.” The233
practitioner can add extra complexity to the model by using the outputs of each hidden234
layer as the inputs for a new fully connected layer. The practitioner has to choose both235
the number of neurons and the depth of the model by choosing the number of hidden236
layers. Once it has been decided that the architecture is deep enough, in the case of binary237
classification such as ours, a final fully connected layer is created yielding a single linear238
combination and the activation of this one is what we consider the “output layer” of the239
network usually normalized between zero and one thanks to the activation function (a240
sigmoid function1is a common choice). The corresponding number in this output layer241
is mapped to a specific class according to a threshold. For example, binary classes code242
their “target” variable as either having values of zero or one. We can say that an output243
layer with a value larger than 0.5 will predict the input belongs to class one and to class244
zero otherwise.245
The question remains on how these networks are actually generalized for large sam-246
ples of images. Consider that a sample of already identified and labeled images, which247
we call our ”target” is compiled in a vector y. Then, we would like to make sure that over-248
all the values of the distinct weights and biases are chosen such that the resulting output249
layer is as close to the target as possible for representative samples. In this case, we would250
like to choose values such that the following distance is minimized via a process called251
“backpropagation” for nobservations in a sample:252
It needs to be noted that deeper networks with hidden layers and many neurons in253
each are capable to make the distance in Equation (1) very small for a sample due to added254
complexity. This however does not come without the risks of making the network attuned255
to only the images fed through the specific sample and incapable of generalizing to others256
from the same population of objects but that were not present in the sampled data. This257
process is called ”overfitting.” Hence, the practicioner needs to be sure to design their258
architecture in a fine balance. The network must be capable to process complex enough259
patterns for classification, but not be so overly attuned to the sample data such that it fails260
classifying data from the same population outside the sample.261
Using open-source optical satellite data from Google Earth (100 x 100 pixels) of tombs264
with known locations and arbitrary patches of land around them, a labelled dataset was265
created with the following labelling scheme:266
0 if tomb present
1 if tomb absent
The dataset is composed of 1212 images with 169 including tombs. Typical observa-267
tions of each case are presented in Figure 8. It is important to note that the distinctive268
shape of the tombs makes them easily distinguishable from other patches of land even in269
low-resolution data.270
In order to verify that the model we fit is a good model, the data is split in two portions,271
one for fitting the model (training data) and one for looking at how well it generalizes272
(testing and validation). The testing and validation data are simply datasets that don’t273
undergo the fitting process. Since the data belongs to the same population as the training274
data does, assessing the goodness of fit of the model in these can give us a good idea of275
Figure 8: Top: Examples of images labelled as tomb absent. Bottom: Examples of images labelled as tomb
how well the model generalizes and it helps identify overfitting.276
The data was split with 75% used for training and 25% used for testing and valida-277
tion. Since the images containing tombs are heavily underrepresented in the dataset,278
augmentation is necessary for training appropriately with multiparameter methods such279
as convolutional neural networks. In this case 655 new images were synthesized from the280
training data with tombs present. The new samples are created by modifying the existing281
ones through randomly zooming, shearing, and performing horizontal flips. Note that282
augmented images are only used during the training stage. Using them for testing or val-283
idation is inappropriate due to their high correlation with the images that they originate284
The CNN utilized for our problem was trained on the augmented data mentioned at the287
beginning of this section. The full summary of the architecture is detailed in Figure 9.288
The CNN was trained in Keras, a Python module which uses Google’s TensorFlow as a289
backend in our case.290
Figure 9: Keras model summary
The architecture shown is relatively simple consisting of 3 convolutional and pooling291
layers with ReLU activations and two fully connected layers before the final activation292
with a sigmoid. The diagram specifies the dimensions of each. For example, the first293
convolutional layer uses 32 filters and outputs a 98 ×98 grid. A natural question not nec-294
essarily explained in the prior sections is what the ”dropout” row means in the diagram.295
Dropout is a regularization technique which disallows certain linear combinations to ex-296
ist at random during the optimization step. This technique helps ”regularize” or penalize297
overfitting. Hence, making sure the model is generalizable.298
Judging the accuracy of the convolutional neural network specified in Section 4.2 requires300
plausible methods for benchmarking. Furthermore, the true metrics of accuracy we are301
interested in are those in the validation data. These would be the ones that would tell302
us how each model works under observations not seen by the training model. As such,303
three models were chosen: a biased random guess, a support vector classifier with a linear304
kernel and a support vector classifier with a radial basis function kernel.305
Random guess is useful as a comparative benchmark since it selects its output by sim-306
ple random chance. In order to make the benchmark tougher, we biased the probabilities307
of classifying an image as containing a tomb to be the proportion of the actual number of308
tombs in the validation set309
Since the shapes of the tombs are simple and easily distinguishable, it stands to rea-310
son that simpler and more tractable classification methods could work as long as they311
allow for flexible boundary classification. Support vector machines with kernels as pro-312
posed by (Boser et al.,1992) work as sensible and powerful alternatives to deep learning313
models. We attempt using two types of kernels in this study, the linear kernel and the314
radial basis function kernel which both allow for different transformations of the data315
pre-classification. Each of these models have their hyperparameters adjusted via 5 -fold316
cross validation.317
We use three measures to compare the predictions made by the classifiers accuracy:318
Precision, Recall and F1score. Definitions below:319
Precision =# of True Positives
# of True Positives + # of False Positives
Recall =# of True Positives
# of True Positives + # of False Negatives
F1score =2
Recall +1
Precision simply gives the rate of correctly classified objects among all classified objects320
with the same label. Recall gives the rate of correctly labeled objects among all actual321
objects with that label. F1score gives a balanced measure of both. All tables and figures322
comparing models in this paper use these measures.323
Table 1and Table 2encapsulate the results obtained from the trained models making324
predictions on the validation set. We can appreciate that for both, images which con-325
tained tombs or those which did not, the CNN performs best. Interestingly, despite the326
fact that SVMs worked under training with an augmented dataset, their performance in327
identifying pictures with tombs was not comparable to that of the neural network. This328
is surprising since the tomb shapes are mostly simple to the naked eye, hence nonlinear329
classification should work well. The reason lies in the likelihood of the SVM models con-330
taining many false positives (objects that are not tombs being identified as such). This331
occurs because other images that might just simply be circular in shape are likely to be332
picked up by the SVM models as tombs . This has been an issue with other detection333
algorithms before e.g. (Caspari et al.,2014) . The big advantage of our architecture relies334
on the quantity of filters used being able to recognize higher subtlety in the patterns of335
the trained dataset that might identify a tomb, beyond just the circular shape. Figure 10336
summarizes both tables and includes a bar for Average/Total, which has a weighted av-337
erage for both classes under the measure. Showing that overall the CNN is the better338
performing model.339
Model Precision Recall F1score
Random Guessing 0.64 0.65 0.65
SVM with linear kernel 0.9 0.96 0.94
SVM with RBF kernel 0.96 0.97 0.97
CNN 0.98 1 0.99
Table 1: Classification metrics for validation data set pictures without tombs.
Model Precision Recall F1score
Random Guessing 0.59 0.58 0.59
SVM with linear kernel 0.29 0.15 0.20
SVM with RBF kernel 0.76 0.67 0.71
CNN 1 0.84 0.91
Table 2: Classification metrics for validation data set pictures with tombs.
Figure 10: Result Summaries
The distinctive shape of the early Iron Age Saka burial mounds and their relatively large341
size make them an ideal training set for machine learning algorithms which can be run342
on open source satellite imagery. Our CNN outperforms other methods and provides343
a valuable approach for the large-scale detection of elite burial mounds in the Eurasian344
steppes. In this way a macro-regional survey of northern Xinjiang and the adjacent ar-345
eas could be conducted in order to assess the spatial distribution of this monument type346
and possibly revise the geographical extent to which Saka-related material culture spread347
through Eastern Central Asia during the first millennium BCE. The method has the clear348
advantage that all analyses can be conducted without the access problems archaeological349
projects in the region usually have to deal with.350
Preliminary satellite imagery analysis has developed into playing a major role in plan-351
ning and implementing archaeological field research (Lasaponara and Masini,2012;Cas-352
pari et al.,2019). But automatic feature detection has yet to become accessible to a wider353
range of researchers in order to be widely applied. A number of attempts have been354
made to connect archaeological surveys with automatic detection of features (Caspari355
et al.,2017;Trier et al.,2009;Trier and Pil,2012), however, it is not commonly used by356
practitioners. Both the complexity of the method which often demands cooperation with357
computer science specialists, and the lack of awareness for the possibility play a role in358
the so far rare application of automatic detection algorithms by archaeological practition-359
ers. The authors do not expect to see a widespread application unless intuitive tools are360
developed for feature selection, algorithm training and visualization of ready-to-use re-361
BENHABILES, H. and TABIA, H. (2016). Convolutional neural network for pottery re-
trieval. Journal of Electronic Imaging,26 (1).
BEWES, J., LOW, A., MO RPH ETT, A., PATE, F. D. and HENNEB ERG , M. (2019). Artificial
intelligence for sex determination of skeletal remains: Application of a deep learning
artificial neural network to human skulls. Journal of Forensic and Legal Medicine,62, 40 –
BOSER, B. E., GUY ON, I. M. and VAPNIK, V. N. (1992). A training algorithm for opti-
mal margin classifiers. In Proceedings of the 5th Annual ACM Workshop on Computational
Learning Theory, ACM Press, pp. 144–152.
J., MAT-GONZLEZ, M. A. and KOUMOUTSAKOS, P. (2019). Automated identification
and deep classification of cut marks on bones and its paleoanthropological implications.
Journal of Computational Science,32, 36 – 43.
CAN, G., ODOBEZ, J.-M. and GATICA-PE REZ , D. (2018). How to tell ancient signs apart?
recognizing and visualizing maya glyphs with cnns. J. Comput. Cult. Herit.,11 (4), 20:1–
CASPARI, G. (2018). Assessing looting from space: The destruction of early iron age buri-
als in northern xinjiang. Heritage,1(2), 320–327.
— (Forthcoming). Quantifying the funerary ritual activity of the late prehistoric southern
kanas region (xinjiang, china). Asian Perspectives.
—, BALZ, T., GAN G, L., WANG, X. and LIAO, M. (2014). Application of hough forests for
the detection of grave mounds in high-resolution satellite imagery.
—, PLETS, G., BALZ, T. and FU, B. (2017). Landscape archaeology in the chinese altai
mountains survey of the heiliutan basin. Archaeological Research in Asia,10 (Complete),
—, SADYKOV, T., BLOCHIN, J., BUESS, M., NIEBERLE, M. and BALZ, T. (2019). Integrating
remote sensing and geophysics for exploring early nomadic funerary architecture in the
siberian valley of the kings. Sensors,19 (14).
CHEN, K. T. and HIEB ERT, F. T. (1995). The late prehistory of xinjiang in relation to its
neighbors. Journal of World Prehistory,9(2), 243–300.
Classification of ceramic shards based on convolutional neural network. 2018 25th IEEE
International Conference on Image Processing (ICIP), pp. 1038–1042.
CLARKE, M. (2008). China’s war on terror in xinjiang: Human security and the causes of
violent uighur separatism. Terrorism and Political Violence,20 (2), 271–301.
DAVIS -KIMBALL, J. (1991). Kazakh/american research project: 1990 field work report.
Middle East Studies Association Bulletin,25 (1), 33–35.
GASS, A. (2011). Early iron age burials in southeastern zhetysu: The geoarchaeological
evidence. Archaeology, Ethnology and Anthropology of Eurasia,39 (3), 57 – 69.
— (2016). Das Siebenstromland Zwischen Bronze- Und Fr¨uheisenzeit: Eine Regionalstudie.
Berlin; Boston; De Gruyter.
GRAHAM, S. (2018). Fleshing out the bones: Studying the human remains trade with
tensorflow and inception. Journal of Computer Applications in Archaeology,1(1), 55–63.
MANN, M., HERVAS, M. A., VILA, L. and ZALLOCCO, M. (2016). ArchAIDE - Archae-
ological Automatic Interpretation and Documentation of cEramics. In C. E. Catalano
and L. D. Luca (eds.), Eurographics Workshop on Graphics and Cultural Heritage, The Eu-
rographics Association.
GUYOT, A., HU BERT-MOY, L. and LORH O, T. (2018). Detecting neolithic burial mounds
from lidar-derived elevation data using a multi-scale approach and machine learning
techniques. Remote Sensing,10 (2).
(2018). Automated classification of archaeological ceramic materials by means of tex-
ture measures. Journal of Archaeological Science: Reports,21, 921 – 928.
HERMOZA, R. and SIPIRAN, I. (2017). 3d reconstruction of incomplete archaeological ob-
jects using a generative adversary network. CoRR,abs/1711.06363.
(1985). Xinjiang xinyuan gongnaisi zhongyangchang shiguan mu (stone cist tomb of
zhongyangchang site, xinyuan, xinjiang). Kaogu yu Wenwu,2, 21–26.
Xinjiang miquan dacaotan faxian shiduimu (the rock-fill tombs discovered at dacaotan,
miquan, xinjiang). Kaogu yu Wenwu,1, 36–38.
Xinjiang xinyuan tiemulike gumu qun (tiemulike cemetery, xinyuan, xinjiang). Kaogu yu
Wenwu,8, 59–66.
ISHITSUKA, K., ISO , S., ONIS HI, K. and MATS UOK A, T. (2018). Object detection in
ground-penetrating radar images using a deep convolutional neural network and im-
age set preparation by migration.
UGEL-BEN NET T, A. and SARGENT, I. (2017). Automated
detection of archaeology in the new forest using deep learning with remote sensor data.
LASAPONARA, R. and MASINI, N. (2012). Satellite remote sensing: a new tool for archaeology.
LECUN, Y., HAFFN ER, P., BOT TOU, L. and BENGIO, Y. (1999). Object recognition with
gradient-based learning, Springer.
LI, Q., ZOU , Q., MA, D., WANG , Q. and WAN G, S. (2018). Dating ancient paintings of
mogao grottoes using deeply learnt visual codes. CoRR,abs/1810.09168.
NAGLER, A. (2009). Grosskurgane im Siebenstromland (Kazachstan), vol. 1. Jahresbericht
2008 des DAI. Arch¨
aologischer Anzeiger 2009.
—, Z, S., H, P. and M, N. (2010). S¨udkasachstan: Kurgane Asy Zaga, Kegen und Zoan Tobe,
Berlin, DAI, chap. Arch¨
aologische Forschungen in Kasachstan, Tadschikistan, Turk-
menistan und Usbekistan., pp. 49–54.
PALANIAPPAN, S. and ADHIKARI, R. (2017). Deep learning the indus script. ArXiv,
PARZINGER, H. (2011). Die fr ¨uhen V¨olker Eurasiens: vom Neolithikum zum Mittelalter. C.H.
phora detection based on a gradient weighted error in a convolution neuronal network.
PHAM, M. and LE F `
EVRE, S. (2018). Buried object detection from b-scan ground penetrat-
ing radar data using faster-rcnn. CoRR,abs/1803.08414.
P’IAN KOV, I. V. (1994). The ethnic history of the sakas. Bulletin of the Asia Institute,8,
ROBERTS, S. R. (2018). The biopolitics of chinas war on terror and the exclusion of the
uyghurs. Critical Asian Studies,50 (2), 232–258.
ROMAN-RANGEL, E. and MARCHAND-MAILLET, S. (2016). Indexing mayan hieroglyphs
with neural codes. In 2016 23rd International Conference on Pattern Recognition (ICPR), pp.
and CLARK, C. M. (2018). Intelligent shipwreck search using autonomous underwater
vehicles. 2018 IEEE International Conference on Robotics and Automation (ICRA), pp. 1–8.
SAMASHEV, Z. (2007). Im Zeichen des goldenen Greifen - K¨onigsgr¨aber der Skythen, Berlin,
Staatliche Museen zu Berlin, chap. Die F ¨
urstengr—”aber des Siebenstromlandes, pp.
SCHOELZ, A. (2018). An embarrassment of riches: Data integration in vr pompeii.
TRAVASSOS, X. L., AVIL A, S. L. and IDA , N. (2018). Artificial neural networks and ma-
chine learning techniques applied to ground penetrating radar: A review. Applied Com-
puting and Informatics.
TRIER,I. D., COWLEY, D. C. and WAL DEL AND , A. U. (2019). Using deep neural networks
on airborne laser scanning data: Results from a case study of semi-automatic mapping
of archaeological topography on arran, scotland. Archaeological Prospection,26 (2), 165–
—, LARSEN , S. Y. and SOLBERG, R. (2009). Automatic detection of circular structures in
high-resolution satellite images of agricultural land. Archaeological Prospection,16 (1),
— and PIL, L. H. (2012). Automatic detection of pit structures in airborne laser scanning
data. Archaeological Prospection,19 (2), 103–121.
(2018). Exploring automated pottery identification [arch-i-scan].
S. and ZAITSEVA, G. (2004). Climate change and the expansion of the scythian culture
after 850 bc: a hypothesis. Journal of Archaeological Science,31 (12), 1735–1742.
WANG, B. (1985). Gudai xinjiang sairen lishi gouchen (a preliminary research of the his-
tory of the ancient saka in xinjiang). Xinjiang Shehui Kexue,1, 48–58, 64.
WANG, H., HE, Z., HU ANG , Y., CHEN, D. and ZHOU, Z. (2017). Bodhisattva head images
modeling style recognition of dazu rock carvings based on deep convolutional network.
Journal of Cultural Heritage,27, 60 – 71.
... Archaeologists and geographers often either manually interpret the aforementioned datasets-high-resolution optical satellite data, and sUAV-derived orthomosaics and DSMs (Caspari 2020;Lunga et al. 2020;Magnini and Bettineschi 2019;Sarasan et al. 2020), or rely on semi-automated mapping procedures like machine-learning algorithms or object-based image analyses (OBIA) with user-defined rulesets (Caspari et al. 2014;Caspari and Crespo 2019;Luo et al. 2019;Magnini and Bettineschi 2019;Orengo and Garcia-Molsosa 2019;Sarasan et al. 2020). Mapping cemeteries and burial mounds have been a particular focus of these studies: Evers and Masters (2018) demonstrated how sUAVs equipped with a near-infrared filter can be used detect both illegal and clandestine burials and assist in forensic work; Caspari (2020) manually mapped burial mounds in the Tuva Republic in southern Siberia using high-resolution (40 cm) panchromatic satellite data; Lunga et al. (2020) used high-resolution (30 and 40 cm) remotely sensed imagery and convolution neural networks to count grave sites in North and South Korea;and Sarasan et al. (2020) used sUAV data and OBIA to map burial mounds across the Plateau in Kyrgyzstan. ...
... As highlighted above, a customized drone equipped with a LiDAR instrument would likely have collected data that would allow for mapping of the entire cemetery, but these systems are costly both in terms of money and, importantly, flight-time as the sensors add weight to the sUAV which decreases the amount of area one can cover on one battery. And although computationally intensive algorithms and methods like those of Caspari et al. 2014;Caspari and Crespo 2019;Luo et al. 2019;Magnini and Bettineschi 2019;Orengo and Garcia-Molsosa 2019;and Sarasan et al. 2020 may have produced similar or slightly better results, a hydrographic analysis using a GIS toolbox was both simple and effective. ...
Full-text available
Estimates suggest that over 15,000 people are buried at East End Cemetery, a historic African American cemetery in Richmond, Virginia, that until recently received no public funding for upkeep. Here, we present a case study analysis and potentially replicable methodology for counting and locating unmarked burial depressions in non-forested areas using a low-cost sUAV (drone) and simple, hydrology-based geographic information systems analyses. Upon visual inspection of 12% of our final 8,000 burial site dataset, we find our dataset is a plausible representation (75% accuracy) of potential grave locations. We hope that the methods presented below can be implemented to assist in reclaiming historically underfunded Black cemeteries across the American South.
... It has experimented with geospatial data/images (satellite, aerial, lidar), texts, categorical tableau data, point clouds, and other datasets. For instance, one can consider some indicative examples such as the work that has been done on bone classification [1], remote sensing archaeology [2][3][4][5][6][7][8][9][10][11][12], geophysical prospection [13][14][15][16][17], detection of objects in paintings [18], classification of pottery [19], and the 3D reconstruction of heritage buildings [20]. The main reason behind this growing trend, which has been noticed in the last five years in all scientific domains, underlies the nuisance generated when dealing with multivariate analysis of high-volume datasets, which are challenging to process and interpret. ...
Full-text available
Ground penetrating radar (GPR) is a well-established technique used in archaeological prospection and it requires a number of specialized routines for signal and image processing to enhance the data acquired and lead towards a better interpretation of them. Computer-aided techniques have advanced the interpretation of GPR data, dealing with a wide range of operations aiming towards locating, imaging, and diagnosis/interpretation. This article will discuss the novel and recent applications of machine learning (ML) and deep learning (DL) techniques, under the artificial intelligence umbrella, for processing GPR measurements within archaeological contexts, and their potential, limitations, and possible future prospects.
... Ref. [30] used a backbone to detect rural dwellings in Indonesia for the ftp-rcnn of ResNet101. Ref. [31] constructed a CNN for detecting royal tombs. The network structure included three convolution layers and two full connection layers, and the results were superior to those of random guessing, SVM with a linear kernel, radial basis function, and SVM with an RBF kernel. ...
Full-text available
As a large-scale irrigation and water conservancy project in ancient times, karez are common in Central Asia and arid regions with a history of thousands of years. Turpan, which is located in the Xinjiang Uyghur Autonomous Region, has the most extensive and concentrated distribution of karez shafts in China. There are tens of thousands of shafts, some of which are in use and are living cultural heritage. According to radiocarbon (14C) dating, some karezs are over 600 years old. The karez is of great significance to the research on geology, hydrology, oasis, climate change, and development history of karez in Turpan. With the development of the population, arable land, industrialization, and urbanization, karez systems are facing the risk of abandonment. Detailed karez distribution mapping or dynamic monitoring data are important for their management or analysis; although there are related methods, due to Turpan’s large desert and “Gobi” environments, field surveys are time- and energy-consuming, and some areas are difficult to access. Precise shaft locations and distribution maps are scarce and often lack georeferencing. The distribution and preservation of karez have not yet been fully grasped. In this study, we evaluated the effectiveness of You Only Look Once version 5 (YOLOv5) in automatically detecting karez in high-resolution images of the Turpan region. We propose post-processing steps to reduce the false karez identified by YOLOv5. Our results demonstrate the feasibility of using YOLOv5 and post-processing techniques to detect karez automatically, and the detected results are sufficient to capture the linear alignment of karez. Target detection based on YOLOv5 and post-processing can greatly improve automatic shaft identification and is therefore useful for the fine mapping of karez. We also applied this method in Shanshan County (for which no detailed mapping data on karez has been obtained before) and successfully detected some karez that had not been archived before. The number of shafts in Turpan is 82,493. Through DBSCAN clustering, it was identified which karez line belonged to which shaft; the number of sections of karez that have been used is 5057, which have a total length of 2387.2 km. The karez line obtained was overlaid with the crop-land data, and the positional relationship between the karez line and the crop land was analyzed. The cultivated area is basically surrounded by karez. Our method can potentially be applied to construct an inventory for all karez shafts globally.
... with the help of Artificial Intelligence (AI) and Machine Learning (ML) [10,30,31,5], it is time to explore the applicability of AI and ML methods in rock art research. ML is a sub-field within AI, where the goal is to learn a model from the given example data. ...
Full-text available
Rock art is globally recognized as significant, yet the resources allocated to the study and exploration of this important form of cultural heritage are often scarce. In areas where numerous rock art sites exist, much of the rock art is unidentified and therefore remains, unrecorded and unresearched. Manually identifying rock art is time-consuming, tedious, and expensive. Therefore, it is necessary to automate many processes in rock art research, which can be accomplished by Machine Learning. Artificial Intelligence (AI) and Machine Learning (ML) can greatly facilitate rock art research in many ways, such as through Object Recognition and Detection, Motif Extraction, Object Reconstruction, Image Knowledge Graphs, and Representations. This article is a reflective work on the future of ML for rock art research. As a proof-of-concept, it presents a machine learning method based on recent advances in deep learning to train a model to identify images with painted rock art (pictograms). The efficacy of the proposed method is shown using data collected from fieldwork in Australia. Furthermore, our proposed method can be used to train models that are specific to the rock art found in different regions. We provide the code and the trained models in the supplementary section.
... Research on automation in archaeological prospection using airborne LiDAR-derived DTM applied different classifiers such as Template Matching [8][9][10], Object-based Image Analysis (OBIA) [11][12][13][14] or Convolutional Neural Networks (CNN). In recent years, CNNs have proven to be powerful tools for detecting different kinds of archaeological objects [15][16][17][18][19][20][21][22]. They repeatedly apply different kernel filters to input images (convolution), that have been known in remote sensing for a long time, e.g., from the field of edge detection. ...
Full-text available
Archaeological research is increasingly embedding individual sites in archaeological contexts and aims at reconstructing entire historical landscapes. In doing so, it benefits from technological developments in the field of archaeological prospection over the last 20 years, including LiDAR-based Digital Terrain Models, special visualizations, and automated site detection. The latter can generate comprehensive datasets with manageable effort that are useful for answering large-scale archaeological research questions. This article presents a highly automated workflow, in which a Convolutional Neural Network is used to detect burial mounds in the proximity of remotely located hollow ways. Detected mounds are then analyzed with respect to their distribution and a possible spatial relation to hollow ways. The detection works well, produces a reasonable number of results, and achieved a precision of at least 77%. The distribution of mounds shows a clear maximum in the radius of 2000-2500 m. This supports future research such as visibility or cost path analysis.
... argue that the authors made some fundamental faults that can easily be addressed. Caspari and Crespo (2019) presented an image classification approach to detect burial mounds on satellite imagery. Their data was split with 75% for training and 25% for testing and validation. ...
Deep learning for automated detection of archaeological sites (objects) on remote sensing data is a highly novel field. The key challenge of this field is in the inherent nature of the objects; they occur in small numbers, are sparsely located and feature a unique pattern on the different remote sensing data modalities. To this extent we identify three main contributions, (1) to include multi-sensor data, (2) to optimise Convolutional Neural Networks (CNNs) for small datasets and, (3) to optimise detection of the sparsely located objects. Our results demonstrate that deep learning can be successfully applied to detect archaeological sites on each of the individual remote sensing images, that our efforts to optimise CNNs for small datasets are successful, and that we have discovered new sites that were missed in a manual data analysis and field survey. We have optimised a workflow for the detection of new archaeological sites. We also share the first large-scale publicly available dataset archaeological image classification and object detection along with benchmarks of the most promising models that we applied in this thesis.
... Machine-learning techniques and algorithms have gradually been included into remote sensing-based archaeological research in recent years, allowing for the automatic discovery of sites and their characteristics. The majority of these applications have used high-resolution satellite imagery [26]. Datasets such as LIDAR [27][28][29] use an active laser sensor that provides an illumination source, making features difficult to detect due to hindrances, such as trees, more evident. ...
Full-text available
This communication article provides a call for unmanned aerial vehicle (UAV) users in archaeology to make imagery data more publicly available while developing a new application to facilitate the use of a common deep learning algorithm (mask region-based convolutional neural network; Mask R-CNN) for instance segmentation. The intent is to provide specialists with a GUI based tool that can apply annotation used for training for neural network models, enable training and development of segmentation models, and allow classification of imagery data to facilitate auto-discovery of features. The tool is generic and can be used for a variety of settings, although the tool was tested using datasets from the United Arab Emirates (UAE), Oman, Iran, Iraq, and Jordan. Current outputs suggest that trained data are able to help identify ruined structures, that is, structures such as burials, exposed building ruins, and other surface features that are in some degraded state. Additionally, qanat(s), or ancient underground channels having surface access holes, and mounded sites, which have distinctive hill-shaped features, are also identified. Other classes are also possible, and the tool helps users make their own training-based approach and feature identification classes. To improve accuracy, we strongly urge greater publication of UAV imagery data by projects using open journal publications and public repositories. This is something done in other fields with UAV data and is now needed in heritage and archaeology. Our tool is provided as part of the outputs given.
The need for a quantitative approach to the morphologic study of ceramics is becoming increasingly evident. Ceramics are the most common material in many archaeological sites and a huge amount of data has accumulated over time. This data can be handled by Machine Learning algorithms which are rapidly gaining popularity in archaeology. Although most approaches can be referred to classification tasks, in this contribution a particular type of Neural Network is proposed for feature extraction from archaeological ceramics. Through this proposed method it is possible to gain a numerical weighted representation of the pottery profile that can be used for multivariate analyses. The case study will focus on regionalisation processes in pottery production between the end of the 2nd millennium BC and the first half of the 1st millennium BC in central Tyrrhenian Italy, a period that saw major transformations in the cultural and socio-political structure of Ancient Italy. The results seem to confirm the regionalisation hypothesis and offer interesting insights into the quantitative study of archaeological ceramics.
Problem formulation is central to recombinant invention, and problem complexity particularly shapes the process and outcome of knowledge recombination. However, research on the antecedents of problem complexity remains limited. This study examines how breakthrough inventions may serve as an important antecedent of the complexity of problems formulated by individual inventors. We propose that a breakthrough invention may facilitate inventors’ appreciation of novel knowledge couplings and improve their overall comprehension of knowledge interdependence for problem formulation, thus increasing problem complexity. We further argue that inventors’ prior search breadth and experimentation strengthen the above effect. By exploiting the unexpected victory of AlphaGo and tracking the questions posted on by developers interested in deep learning, we find empirical evidence that supports our hypotheses. Because formulating complex problems may result in novel and valuable inventions, it is helpful to understand what drives inventors to formulate problems of higher complexity. In this study, we propose that breakthrough inventions can serve as an important antecedent of problem complexity. In the context of deep learning, we find that developers formulate problems of higher complexity after the unexpected victory of AlphaGo, a widely acknowledged inventive breakthrough. Further, the above effect is more pronounced when developers have broad search experience in the past or engage frequently in experimentation. The insights of this study are not only relevant to managerial practices within the dynamic field of deep learning but also generalizable to broader technological contexts with breakthrough inventions.
Predictive modeling approaches and techniques in archaeological research and cultural heritage management have experienced continuous developments. In this Chapter, a more specific prospect of past and present modeling approaches in archaeology is provided. Recent and concrete applications and case studies applying statistical learning are introduced, ranging from a global to a local (Swiss) perspective. The incorporation of uncertainty and vagueness inherent to the data and its delivering in Archaeological Predictive Maps is further tackled and summarized. Finally, more complex, non-linear machine learning and data mining approaches are described, with particular emphasis on the Random Forest algorithm as fundamental part of the methodological procedure developed in this study.
Full-text available
This article analyses the architecture of the Early Iron Age royal burial mound Tunnug 1 in the “Siberian Valley of the Kings” in Tuva Republic, Russia. This large monument is paramount for the archaeological exploration of the early Scythian period in the Eurasian steppes, but environmental parameters make research on site difficult and require the application of a diversity of methods. We thus integrate WorldView-2 and ALOS-2 remote sensing data, geoelectric resistivity and geomagnetic survey results, photogrammetry-based DEMs, and ortho-photographs, as well as excavation in order to explore different aspects of the funerary architecture of this early nomadic monument. We find that the large royal tomb comprises of a complex internal structure of radial features and chambers, and a rich periphery of funerary and ritual structures. Geomagnetometry proved to be the most effective approach for a detailed evaluation of the funerary architecture in our case. The parallel application of several surveying methods is advisable since dataset comparison is indispensable for providing context.
Full-text available
A deep learning artificial neural network was adapted to the task of sex determination of skeletal remains. The neural network was trained on images of 900 skulls virtually reconstructed from hospital CT scans. When tested on previously unseen images of skulls, the artificial neural network showed 95% accuracy at sex determination. Artificial intelligence methods require no significant expertise to implement once trained, are rapid to use, and have the potential to eliminate human bias from sex estimation of skeletal remains.
Full-text available
Ground-penetrating radar allows the acquisition of many images for investigation of the pavement interior and shallow geological structures. Accordingly, an efficient methodology of detecting objects, such as pipes, reinforcing steel bars, and internal voids, in ground-penetrating radar images is an emerging technology. In this paper, we propose using a deep convolutional neural network to detect characteristic hyperbolic signatures from embedded objects. As a first step, we developed a migration-based method to collect many training data and created 53510 categorized images. We then examined the accuracy of the deep convolutional neural network in detecting the signatures. The accuracy of the classification was 0.945 (94.5%)–0.979 (97.9%) when using several thousands of training images and was much better than the accuracy of the conventional neural network approach. Our results demonstrate the effectiveness of the deep convolutional neural network in detecting characteristic events in ground-penetrating radar images.
Full-text available
Burial mounds (kurgans) of the Early Iron Age in the steppe zones of Central Asia have long been the target of severe looting activities. Protection of these monuments in remote areas is difficult since accurate mapping is rarely available. We map an area in northern Xinjiang using a combination of high-resolution optical data and on-ground survey to establish a quantitative and qualitative assessment of looting. We find that at least 74.5% of burial mounds are looted or otherwise destroyed. Due to the large number of visibly impacted burial mounds, it becomes clear that the bulk of cultural heritage of the Early Iron Age in this area is under threat. The looting, however, continues until present day. Rescue excavation of potentially untouched burials in the area is advisable.
Full-text available
Ground Penetrating Radar is a multidisciplinary Nondestructive Evaluation technique that requires knowledge of electromagnetic wave propagation, material properties and antenna theory. Under some circumstances this tool may require auxiliary algorithms to improve the interpretation of the collected data. Detection, location and definition of target's geometrical and physical properties with a low false alarm rate are the objectives of these signal post-processing methods. Basic approaches are focused in the first two objectives while more robust and complex techniques deal with all objectives at once. This work reviews the use of Artificial Neural Networks and Machine Learning for data interpretation of Ground Penetrating Radar surveys. We show that these computational techniques have progressed GPR forward from locating and testing to imaging and diagnosis approaches.
Identifying peaks in anthropogenic activity in a landscape is an important starting point for understanding past social dynamics in the longue durée. Through intensive surveys and remote sensing surveys of the Heiliutan Basin (Heiliutan Dacaoyuan 黑流滩大草原) in the southern Kanas Region (Kanasi 喀纳斯), Xinjiang, China, a high-resolution dataset for over 4000 years of material culture is established. The complete coverage of the area of interest allows for the quantification of ritual funerary activity based on the number of constructed monuments per century. The data show that the intensity of ritual funerary activity was very low and only left marginal traces in the landscape from the Eneolithic Age to the Late Bronze Age. During the Early Iron Age (ca. 850-200 B.C.E.), the basin became a center for construction of burials for social elites of nomadic tribes and the area was rapidly transformed into a landscape of the dead. The Late Iron Age (starting ≈200 B.C.E.) saw a decline of ritual funerary activities in the basin as it became an unimportant side scene to the cultural developments of the wider region.
The identification of cut marks and other bone surface modifications (BSM) provides evidence for the emergence of meat-eating in human evolution. This most crucial part of taphonomic analysis of the archaeological human record has been controversial due to highly subjective interpretations of BSM. Here, we use a sample of 79 trampling and cut marks to compare the accuracy in mark identification on bones by human experts and computer trained algorithms. We demonstrate that deep convolutional neural networks (DCNN) and support vector machines (SVM) can recognize marks with accuracy that far exceeds that of human experts. Automated recognition and analysis of BSM using DCNN can achieve an accuracy of 91% of correct identification of cut and trampling marks versus a much lower accuracy rate (63%) obtained by trained human experts. This success underscores the capability of machine learning algorithms to help resolve controversies in taphonomic research and, more specifically, in the study of bone surface modifications. We envision that the proposed methods can help resolve on-going controversies on the earliest human meat-eating behaviors in Africa and other issues such as the earliest occupation of America.
Thanks to the digital preservation of cultural heritage materials, multimedia tools (e.g., based on automatic visual processing) considerably ease the work of scholars in the humanities and help them to perform quantitative analysis of their data. In this context, this article assesses three different Convolutional Neural Network (CNN) architectures along with three learning approaches to train them for hieroglyph classification, which is a very challenging task due to the limited availability of segmented ancient Maya glyphs. More precisely, the first approach, the baseline, relies on pretrained networks as feature extractor. The second one investigates a transfer learning method by fine-tuning a pretrained network for our glyph classification task. The third approach considers directly training networks from scratch with our glyph data. The merits of three different network architectures are compared: a generic sequential model (i.e., LeNet), a sketch-specific sequential network (i.e., Sketch-a-Net), and the recent Residual Networks. The sketch-specific model trained from scratch outperforms other models and training strategies. Even for a challenging 150-class classification task, this model achieves 70.3% average accuracy and proves itself promising in case of a small amount of cultural heritage shape data. Furthermore, we visualize the discriminative parts of glyphs with the recent Grad-CAM method, and demonstrate that the discriminative parts learned by the model agree, in general, with the expert annotation of the glyph specificity (diagnostic features). Finally, as a step toward systematic evaluation of these visualizations, we conduct a perceptual crowdsourcing study. Specifically, we analyze the interpretability of the representations from Sketch-a-Net and ResNet-50. Overall, our article takes two important steps toward providing tools to scholars in the digital humanities: increased performance for automation and improved interpretability of algorithms.