Content uploaded by Lawrence Roberts
Author content
All content in this area was uploaded by Lawrence Roberts on Nov 19, 2014
Content may be subject to copyright.
Machine Perception of Three-Dimensional Solids
Abstract
by Lawrence Gilman Roberts.
... The basic act of machine recognition of a real-world object follows a two-step process. In the first step, the inner characterization of the real-world scenario is expressed in a form or model, and in the second step obtained internal form of the scene is compared with other known forms of known objects [2]. The difficulties encountered by machines in recognizing 3D real-world objects have been explained below. ...
... The history of object recognition is ancient and started its journey in the 60s [2]. Plenty of research works have been carried out and are still being carried out in present times as matching machine visibility just like its human counterpart remains unsolved. ...
Since 3D models and printing have become more and more widespread, 3D object recognition has gained appeal. It might be challenging to locate the needed object in a large database of 3D objects of different kinds. Several methods have been put forth in the past to effectively identify and detect 3D objects based on their shape data. These techniques describe 3D objects in a high-dimensional feature plane either by using geometric shape features or by creating a 3D shape model from a 2D shape model. One of the disadvantages of 3D descriptors is that their high memory requirements make recognition speed a major difficulty. We provide a unique 3D object search method based on 3D object identification to get around the problem. The process reads the 3D vertices from a CAD file, samples the triangular meshes from below, locates the center of each triangle, and then projects the points onto the front and top views of the YZ and XY planes, respectively, in two dimensions. Ten clusters are formed from the points in each plane, and the feature descriptor for the three-dimensional object is derived from the cluster centers. The shape is indexed using this description. In order to extract the descriptor from CAD files, the system looks for objects in a designated directory. A feed-forward neural network is then used to classify the descriptor. We demonstrate that the proposed method outperforms the existing shape descriptor-based methods in terms of speed 139 ms, classification accuracy of 98.3%, and search time.
... [28] smooths the image by reducing noise and details, resulting in a so er, blurrier appearance. In contrast, Roberts cross filter [29] suppresses overall content, emphasizing details and textures. ...
Today, it is easier than ever to manipulate images for unethical purposes. This practice is therefore increasingly prevalent in social networks and advertising. Malicious users can for instance generate convincing deep fakes in a few seconds to lure a naive public. Alternatively, they can also communicate secretly hidding illegal information into images. Such abilities raise significant security concerns regarding misinformation and clandestine communications. The Forensics community thus actively collaborates with Law Enforcement Agencies worldwide to detect image manipulations.
The most effective methodologies for image forensics rely heavily on convolutional neural networks meticulously trained on controlled databases. These databases are actually curated by researchers to serve specific purposes, resulting in a great disparity from the real-world datasets encountered by forensic practitioners. This data shift addresses a clear challenge for practitioners, hindering the effectiveness of standardized forensics models when applied in practical situations.
Through this thesis, we aim to improve the efficiency of forensics models in practical settings, designing strategies to mitigate the impact of data shift. It starts by exploring literature on out-of-distribution generalization to find existing strategies already helping practitioners to make efficient forensic detectors in practice. Two main frameworks notably hold promise: the implementation of models inherently able to learn how to generalize on images coming from a new database, or the construction of a representative training base allowing forensics models to generalize effectively on scrutinized images. Both frameworks are covered in this manuscript.
When faced with many unlabeled images to examine, domain adaptation strategies matching training and testing bases in latent spaces are designed to mitigate data shifts encountered by practitioners. Unfortunately, these strategies often fail in practice despite their theoretical efficiency, because they assume that scrutinized images are balanced, an assumption unrealistic for forensic analysts, as suspects might be for instance entirely innocent. Additionally, such strategies are tested typically assuming that an appropriate training set has been chosen from the beginning, to facilitate adaptation on the new distribution..
Trying to generalize on a few images is more realistic but much more difficult by essence. We precisely deal with this scenario in the second part of this thesis, gaining a deeper understanding of data shifts in digital image forensics. Exploring the influence of traditional processing operations on the statistical properties of developed images, we formulate several strategies to select or create training databases relevant for a small amount of images under scrutiny.
Our final contribution is a framework leveraging statistical properties of images to build relevant training sets for any testing set in image manipulation detection. This approach improves by far the generalization of classical steganalysis detectors on practical sets encountered by forensic analyst and can be extended to other forensic contexts.
... On 1963, Lawrence Gilman Roberts, known as Larry Roberts, presented his thesis "Machine Perception of Three-Dimensional Solids" where he presented methods to reconstruct 3D objects from 2D images, paving the path for 3D Computer Vision. He also introduced the idea of edge detection to identify the boundaries of objects in an image as well as mathematical models, which all contributed to future developments of Computer Vision technologies (Roberts, 1963). ...
The field of Computer Vision, a pivotal subdomain of Artificial Intelligence (AI), has seen extraordinary advancements since its emergence in the 1960s. This paper examines the historical development of Computer Vision technologies, tracing the journey from early foundational models, such as Frank Rosenblatt’s Perceptron, to contemporary breakthroughs driven by Deep Learning. Key milestones are explored, including the development of algorithms like Scale-Invariant Feature Transform (SIFT), Viola-Jones for face detection, and Eigenfaces, which paved the way for modern solutions such as Convolutional Neural Networks (CNNs), YOLO and FaceNet. The paper highlights the evolution of face detection and recognition techniques, contrasting traditional methods with the transformative capabilities of Deep Learning-driven approaches. Additionally, we analyze the growing computational demands of modern algorithms, discussing the trade-offs between accuracy and efficiency and their implications for practical applications. This study underscores the rapid progression of Computer Vision, its challenges, and its role as a cornerstone in shaping the future of Artificial Intelligence.
... Object Reconstruction. Since the foundational work by Roberts [54], numerous methods have been developed to learn cues for deriving 3D object structures, thereby bridging the gap between 2D perception and the 3D world. These methods typically involve an image encoder network that processes the input image of a single object, capturing its features. ...
We present a novel diffusion-based approach for coherent 3D scene reconstruction from a single RGB image. Our method utilizes an image-conditioned 3D scene diffusion model to simultaneously denoise the 3D poses and geometries of all objects within the scene. Motivated by the ill-posed nature of the task and to obtain consistent scene reconstruction results, we learn a generative scene prior by conditioning on all scene objects simultaneously to capture the scene context and by allowing the model to learn inter-object relationships throughout the diffusion process. We further propose an efficient surface alignment loss to facilitate training even in the absence of full ground-truth annotation, which is common in publicly available datasets. This loss leverages an expressive shape representation, which enables direct point sampling from intermediate shape predictions. By framing the task of single RGB image 3D scene reconstruction as a conditional diffusion process, our approach surpasses current state-of-the-art methods, achieving a 12.04% improvement in AP3D on SUN RGB-D and a 13.43% increase in F-Score on Pix3D.
... This reconstructive framework is also known as analysisby-synthesis (Grenander, 1978), or vision-as-inverse-graphics (VIG) (Kulkarni et al., 2015;Moreno et al., 2016). It can be traced back to the early days of "blocks world" research in the 1960s (Roberts, 1963). Other early work in this vein includes the VISIONS system of Hanson and Riseman (1978), and the system of Ohta et al. (1978) for outdoor scene understanding. ...
This position paper argues for the use of structured generative models (SGMs) for the understanding of static scenes. This requires the reconstruction of a 3D scene from an input image (or a set of multi-view images), whereby the contents of the image(s) are causally explained in terms of models of instantiated objects, each with their own type, shape, appearance and pose, along with global variables like scene lighting and camera parameters. This approach also requires scene models which account for the co-occurrences and inter-relationships of objects in a scene. The SGM approach has the merits that it is compositional and generative, which lead to interpretability and editability. To pursue the SGM agenda, we need models for objects and scenes, and approaches to carry out inference. We first review models for objects, which include “things” (object categories that have a well defined shape), and “stuff” (categories which have amorphous spatial extent). We then move on to review scene models which describe the inter-relationships of objects. Perhaps the most challenging problem for SGMs is inference of the objects, lighting and camera parameters, and scene inter-relationships from input consisting of a single or multiple images. We conclude with a discussion of issues that need addressing to advance the SGM agenda.
... Contour Detection: Early approaches to contour detection relied on local gradient measurements in an image [20], [21], [22]. These simple edge detectors operate by applying local derivative filters on grayscale images. ...
We present Convolutional Oriented Boundaries (COB), which produces multiscale oriented contours and region hierarchies starting from generic image classification Convolutional Neural Networks (CNNs). COB is computationally efficient, because it requires a single CNN forward pass for multi-scale contour detection and it uses a novel sparse boundary representation for hierarchical segmentation; it gives a significant leap in performance over the state-of-the-art, and it generalizes very well to unseen categories and datasets. Particularly, we show that learning to estimate not only contour strength but also orientation provides more accurate results. We perform extensive experiments for low-level applications on BSDS, PASCAL Context, PASCAL Segmentation, and NYUD to evaluate boundary detection performance, showing that COB provides state-of-the-art contours and region hierarchies in all datasets. We also evaluate COB on high-level tasks when coupled with multiple pipelines for object proposals, semantic contours, semantic segmentation, and object detection on MS-COCO, SBD, and PASCAL; showing that COB also improves the results for all tasks.
In this study, we propose a new edge detection method. First, we redefine the concept of “Edge Region” and mention that the edge-narrowness principle for the edge detection could be originally introduced based on the “Edge Region”. Next, contour lines of the object are discussed, and the principle of the “Edge Region” expression be explained based on them. Based on the above considerations, we propose a new method called the Narrowness Edge Detection (NED) method. Experiments showed the behavior of the algorithm of the proposed method. The position of the proposed method was clarified by comparing it with methods based on the local contrast principle such as Sobel operator.
Edge and contour detection plays critical roles in computer vision and image processing, with extensive applications in advanced tasks including object recognition, shape matching, visual saliency, image segmentation, and inpainting. In recent decades, this field has attracted significant attention, leading to the development of numerous sophisticated methods that approximate human visual performance. Despite these advances, notable gaps remain. This review offers a comprehensive analysis of representative techniques, categorizing them into traditional and learning-based approaches, and examines their strengths and limitations to identify the underlying reasons for these gaps. Traditional methods are further divided into four sub-categories: local pattern, edge grouping, active contour, and bio-inspired techniques, with a specific emphasis on the promising potential of bio-inspired methods. Learning-based approaches, on the other hand, are classified into two types: classical learning, which typically relies on handcrafted features designed from empirical knowledge, and deep learning, which autonomously extracts features from large-scale datasets without human intervention. Additionally, benchmarks and evaluation metrics related to edge and contour detection are discussed, with potential issues identified within these frameworks. A quantitative assessment of representative methods is conducted across three popular benchmarks. Lastly, challenges and future prospects in edge and contour detection are explored, focusing on five key aspects: model architecture, learning strategies, feature extraction and fusion, method integration, and cross-domain applications. These considerations aim to bridge the gap with human visual perception. Overall, this work is expected to benefit researchers and advance progress in the field.
ResearchGate has not been able to resolve any references for this publication.