![The OCC model (source: [2])](https://www.researchgate.net/profile/Adam-Wojciechowski-3/publication/324263019/figure/fig47/AS:612515193053191@1523046410697/The-OCC-model-source-2_Q320.jpg)
Content uploaded by Adam Wojciechowski
Author content
All content in this area was uploaded by Adam Wojciechowski on Mar 25, 2020
Content may be subject to copyright.
A preview of the PDF is not available
Computer Game Innovations 2017
Abstract and Figures
The "Computer Game Innovations" series is an international forum designed to enable the exchange of knowledge and expertise in the field of video game development. Comprising both academic research and industrial needs, the series aims at advancing innovative industry-academia collaboration. The monograph provides a unique set of articles presenting original research conducted in the leading academic centres which specialise in video games education. The goal of the publication is, among others, to enhance networking opportunities for industry and university representatives seeking to form R&D partnerships. This publication covers the key focus areas specified in the GAMEINN sectoral programme supported by the National Centre for Research and Development.
Figures - uploaded by Adam Wojciechowski
Author content
All figure content in this area was uploaded by Adam Wojciechowski
Content may be subject to copyright.
ResearchGate has not been able to resolve any citations for this publication.
Recent work has shown that optical flow estimation can be formulated as a supervised learning problem. Moreover, convolutional networks have been successfully applied to this task. However, supervised flow learning is obfuscated by the shortage of labeled training data. As a consequence, existing methods have to turn to large synthetic datasets for easily computer generated ground truth. In this work, we explore if a deep network for flow estimation can be trained without supervision. Using image warping by the estimated flow, we devise a simple yet effective unsupervised method for learning optical flow, by directly minimizing photometric consistency. We demonstrate that a flow network can be trained from end-to-end using our unsupervised scheme. In some cases, our results come tantalizingly close to the performance of methods trained with full supervision.
Gender recognition, across different races and regardless of age, is becoming an increasingly important technology in the domains of marketing, human-computer interaction and security. Most state-of-the-art systems consider either highly constrained conditions or relatively large databases. In either case, often not enough attention is paid to cross-racial age-invariant applications. This paper proposes a~method of hybrid classification, which performs well even with a small training set. The design of the classifier enables the construction of reliable decision boundaries insensitive to an aging model as well as to race variation. For a training set consisting of one hundred images, the proposed method reached an accuracy level of 90%, whereas the best method known from the literature, tested under the restrictions imposed on the database, achieved only 78% accuracy.
The OpenMVG C++ library provides a vast collection of multiple-view geometry tools and algorithms to spread the usage of computer vision and structure-from-motion techniques. Close to the state-of-the-art in its domain, it provides an easy access to common tools used in 3D reconstruction from images. Following the credo “Keep it simple, keep it maintainable” the library is designed as a modular collection of algorithms, libraries and binaries that can be used independently or as bricks to build larger systems. Thanks to its strict test driven development, the library is packaged with unit-test code samples that make the library easy to learn, modify and use. Since its first release in 2013 under the MPL2 license, OpenMVG has gathered an active community of users and contributors from many fields, spanning hobbyists, students, computer vision experts, and industry members.
The objective of this course is to provide newcomers to Human-Computer Interaction (HCI) with an introduction and overview of the field. Attendees often include practitioners without a formal education in HCI, and those teaching HCI for the first time. This course includes content on theory, cognition, design, evaluation, and user diversity.
We trained a large, deep convolutional neural network to classify the 1.2 million high-resolution images in the ImageNet LSVRC-2010 contest into the 1000 dif- ferent classes. On the test data, we achieved top-1 and top-5 error rates of 37.5% and 17.0% which is considerably better than the previous state-of-the-art. The neural network, which has 60 million parameters and 650,000 neurons, consists of five convolutional layers, some of which are followed by max-pooling layers, and three fully-connected layers with a final 1000-way softmax. To make training faster, we used non-saturating neurons and a very efficient GPU implemen- tation of the convolution operation. To reduce overfitting in the fully-connected layers we employed a recently-developed regularization method called dropout that proved to be very effective. We also entered a variant of this model in the ILSVRC-2012 competition and achieved a winning top-5 test error rate of 15.3%, compared to 26.2% achieved by the second-best entry
An increasing number of our technological interactions are mediated through virtually embodied characters and software agents powered by machine learning. To understand how users relate to and evaluate these types of interfaces, we designed a Wizard of Oz prototype of an embodied agent in Minecraft that learns from users' actions, and conducted a user study with 18 school-aged Minecraft players. We categorised nine main ways users spontaneously attempted to interact with and teach the agent: four using game controls, and five using natural language text input. This study lays groundwork for a better understanding of human interaction with learning agents in virtual worlds.
You're being asked to quantify your usability improvements with statistics. But even with a background in statistics, you are hesitant to statistically analyze their data, as they are often unsure which statistical tests to use and have trouble defending the use of small test sample sizes. The book is about providing a practical guide on how to solve common quantitative problems arising in usability testing with statistics. It addresses common questions you face every day such as: Is the current product more usable than our competition? Can we be sure at least 70% of users can complete the task on the 1st attempt? How long will it take users to purchase products on the website? This book shows you which test to use, and how provide a foundation for both the statistical theory and best practices in applying them. The authors draw on decades of statistical literature from Human Factors, Industrial Engineering and Psychology, as well as their own published research to provide the best solutions. They provide both concrete solutions (excel formula, links to their own web-calculators) along with an engaging discussion about the statistical reasons for why the tests work, and how to effectively communicate the results. *Provides practical guidance on solving usability testing problems with statistics for any project, including those using Six Sigma practices *Show practitioners which test to use, why they work, best practices in application, along with easy-to-use excel formulas and web-calculators for analyzing data *Recommends ways for practitioners to communicate results to stakeholders in plain English. © 2012 Jeff Sauro and James R. Lewis Published by Elsevier Inc. All rights reserved.
This paper presents a method of improving the estimation of distances between an autonomous harvesting robot and plants with ripe fruits by using the vision system based on five cameras. The system is called Equal Baseline Multiple Camera Set (EBMCS). EBMCS has some features of a camera matrix and a camera array. EBMCS is regarded as a set of stereo cameras for estimating distances by obtaining disparity maps and depth maps. This paper introduces Exceptions Excluding Merging Method (EEMM) which makes it possible to improve the quality of disparity maps by integrating maps acquired from individual stereo cameras included in EBMCS. The method was tested with eight different stereo matching algorithms including Efficient Large-scale Stereo Matching (ELAS), algorithms implemented in the OpenCV library and algorithms available in Middlebury Stereo Vision Page. Experiments were performed on input data sets which contained images of strawberry plants, cherry trees and redcurrant plants. The bad matching pixels (BMP) metric was used for measuring the error rate in disparity maps used in the distance estimation. The results of experiments showed that, on average, the EEMM merging method used with EBMCS consisting of five cameras reduces the error rate of the distance estimation by 26.55% in comparison to results obtained from stereoscopy based on a single stereo camera. The best results were acquired by using with five cameras a stereo matching algorithm based on a graph cut.
In this paper, we present an application of rotational gain to horizontal panning for viewing 360 degree videos. Rotational gain refers to the ratio between the rotation velocity in the virtual (video) space to that of the physical, and it allows the user to rotate one's head less quickly than actually needed (without the user noticing such adjustment to some degree). As such it can bring about convenience, and less physical movement, fatigue and possibly even sickness. We implemented a 360 video panning system with both a constant and dynamic gain, and compared user behavior and subjective usability. Our pilot study showed promising results in that with the proper gain value and control scheme, the user will unknowingly use less physical movement than needed, yet maintain reasonable spatial understanding with higher usability.