Papers by Sudhanshu Semwal
SIGGRAPH Asia 2021 Technical Communications, 2021
In this paper we propose the Advanced Deep Learning Graphics Pipeline (ADLGP). ADLGP is a novel a... more In this paper we propose the Advanced Deep Learning Graphics Pipeline (ADLGP). ADLGP is a novel approach that uses existing deep learning architectures to convert scene data into rendered images. Our goal of generating frames from semantic data has produced successful renderings with similar structures and composition as target frames. We demonstrate the success of ADLGP with side-by-side comparisons of frames generated through standard rendering procedures. We assert that a fully implemented ADLGP framework would reduce the time spent in visualizing 3D environments, and help selectively offload the requirements of the current graphics rendering pipeline. CCS CONCEPTS • Computing methodologies → Rendering; Image processing; Machine learning algorithms.
The graphics rendering pipeline is key to generating realistic images, and is a vital process of ... more The graphics rendering pipeline is key to generating realistic images, and is a vital process of computational design,modeling, games, and animation. Perhaps the largest limiting factor of rendering is time; the processing requiredfor each pixel inevitably slows down rendering and produces a bottleneck which limits the speed and potential ofthe rendering pipeline. We applied deep generative networks to the complex problem of rendering an animated 3Dscene. Novel datasets of annotated image blocks were used to train an existing attentional generative adversarialnetwork to output renders of a 3D environment. The annotated Caltech-UCSD Birds-200-2011 dataset served asa baseline for comparison of loss and image quality. While our work does not yet generate production qualityrenders, we show how our method of using existing machine learning architectures and novel text and imageprocessing has the potential to produce a functioning deep rendering framework.
In this paper, we summarize the present methods available for providing user interfaces for disab... more In this paper, we summarize the present methods available for providing user interfaces for disability applications. Section 508 of the Rehabilitation Act of 1973 has contributed to providing accommodation to people with disability, and is generally considered a step in the right direction. As many modalities of interaction have become a focus, we propose multi-channel customized interfaces, where a person can take many types of inputs and map some portions of each of them to the output channels available, emerging as a sound design principle. Uniquely customizing interfaces provides an opportunity for all, especially people with disabilities. In addition, the possibility of creating user interfaces for special cases and on demand helps to provide inclusive computing.
International Conferences Interfaces and Human Computer Interaction 2019; Game and Entertainment Technologies 2019; and Computer Graphics, Visualization, Computer Vision and Image Processing 2019, 2019
The Gamma Knife radiosurgery treatment involves geometrically mapping for covering the tumor area... more The Gamma Knife radiosurgery treatment involves geometrically mapping for covering the tumor area in three dimensions so that the gamma rays can provide precise treatment to the affected area. The entire process has to be accurate in order to prevent the exposure of the gamma rays to the neighboring healthy tissues. When the tumor shape is irregular, it is difficult to ensure the coverage of the tumor region. We present, in this paper, a new approach to fit spheres in a bounded region based on maximum distance. Our proposed method for automating the gamma knife surgery planning is by using a geometrically-based Euclidean sphere packing optimization method. We implement our algorithm in Slicer3D using Python libraries. Experiments show that our algorithm produces reasonable dense packing of different irregular shapes.
Word Sense Disambiguation (WSD) is a task which aims to identify the meaning of a word given its ... more Word Sense Disambiguation (WSD) is a task which aims to identify the meaning of a word given its context. This problem has been investigated and analyzed in depth in English. However, work in Arabic has been limited despite the fact that there are half a billion native Arabic speakers. In this work, we present multiple approaches for the problem of WSD in Arabic, inspired by recent developments and successes in learning word embeddings with approaches such as GloVe, and Word2vec. The primary shortcoming of word embeddings is the single vector representation of a word’s meaning, although many words are polysemous. Our first contribution in this work is to computationally obtain an embedding for each sense, using an Arabic WordNet (AWN) to overcome the problem of WSD. We also compute word semantic similarity giving thought to multiple Arabic stemming algorithms. We tackle WSD from a different point of view that solves WSD by restoring Arabic diacritics to reduce the complexity of word...
Many parking lot facilities suffer from capacity over-loads and many times there are no monitorin... more Many parking lot facilities suffer from capacity over-loads and many times there are no monitoring tools toprovide feedback. As a consequence, the people, want-ing to park, become frustrated as there is considerablyloss of time. In this paper, we present a novel proto-type of an automatic parking-lot analysis platform us-ing image-based machine learning to (a) guide a droneautonomously; and (b) to process useful information tobe handled into a smartphone application to communi-cate with the parking lot users.We have collected a reasonable amount of test imagesto build a classification model using Convolutional neu-ronal networks (CNNs) to classify parking lot images,and build different object detection models to identifyfree and occupied parking spots. Those models havebeen exported to the back-end module of our platformso it can control the drone and record the computed in-formation to its database. In addition, we have imple-mented an iOS application that requests and displaysthe ...
This paper proposes a web-based terminal for teaching the cloud. This method allows students and ... more This paper proposes a web-based terminal for teaching the cloud. This method allows students and trainees to learn managing the cloud in a useful and secure manner. This work promotes an appreciate educational environment by enabling students to practice the knowledge practically, teaching them the ethics of using educational tools, and reducing the learning distraction.It is also considered a crossover applications by using it in education, training, and actual management. The web-based terminal was implemented and tested. The results showed that our implementation improves Xen (the cloud example we used)usability and accessibility by allowing users to learn managing the cloud while obviating the use of SSH-clients and reducing the risk of giving everyone a Linux superuser account. Our methodology is general and applicable to other virtualization and cloud solutions.
Bounding Volume Hierarchies (BVHs) and k-d trees have been used to create interactive ray tracing... more Bounding Volume Hierarchies (BVHs) and k-d trees have been used to create interactive ray tracing. Ray tracing dynamic scenes using nVidia’s R © OptiX TM has already provided thirty to sixty frames per second or better. In this paper, we implement space partitioning methods, such as grid method and the proximity clouds (PCs), on multiple GPUs. Our motivation is to investigate the use of such methods for medical applications, because there is direct one to one correspondence between 3D voxels used in space partitioning methods and 3D voxels in the volume data. In the past, proximity Clouds have worked well on static scenes, but object movement forces recalculation of the scene and some preprocessing cost. This paper investigates parallelizing these techniques on the GPU to determine the feasibility of dynamic scenes using them. Our scenes are made of spheres instead of volume data because at this time we do not know of any technique that can generate dynamically changing volume data....
Scalable Spherical Harmonics Hierarchies (SSPHH) is a real-time rendering solution to the global ... more Scalable Spherical Harmonics Hierarchies (SSPHH) is a real-time rendering solution to the global illumination problem. Our novel method is a system of components that enables the computation of light probes and the conversion to spherical harmonics coefficients, which are used as anisotropic time-varying light sources we call Spherical Harmonics Lights (SPHLs). The SPHLs encode irradiance information that can be used for imagebased lighting. Our approach focuses on reconstructing scene lighting using diffuse illumination but is flexible to allow specular details. Furthermore, we consider the light transport from neighboring SPHLs by computing a transfer coefficient that estimates how much light from one probe is visible at another. SPHLs can be used for physically-based lighting using rendering methods similar to point lights and shadow maps. We created a reproducible testing methodology to compare our images with those of a commercial path tracer by automatically generating the app...
Proceedings of the Future Technologies Conference (FTC) 2021, Volume 1, 2021
2016 International Conference on Collaboration Technologies and Systems (CTS), 2016
What if a person's life choices affected their game experience? What if it affected all of th... more What if a person's life choices affected their game experience? What if it affected all of their games? Or, what if the game experience affected all the life choices? That is the central idea behind this paper, we are proposing an integration and collaboration between the real andthe virtual world at an individual level, affecting multitude of digital world that the average gamer inhabit based on their life choices. In our implementation, a parent can reward or encourage an active lifestyle such as walking, hiking or sports in the real world by controlling access to their kid's digital world. A user of our implementation looking for motivation for a more physically active lifestyle could work towards their fitness goal while being rewarded for their physical effort by increasing the amount of time they can play a video game of their choice. We use Unity3Dâ„¢ game engine to show the feasibility of our implementation. We hope that our system will be considered an innovative way of rewarding good behavior in real-life. Our idea also can be considered an example of novel gameplay mechanics.
Real-Time Imaging, 1999
... 1999 Academic Press Sudhanshu K. Semwal and Michael J. Parker Department of Computer Science,... more ... 1999 Academic Press Sudhanshu K. Semwal and Michael J. Parker Department of Computer Science, University ... Edmund Burke at the University of Colorado at Colorado Springs, and Dr Jeffrey Broker ... Kelso, JAS Pandya, AS (1991) Dynamic pattern genera tion and recognition ...
2015 International Conference on Collaboration Technologies and Systems (CTS), 2015
The relationship between music and emotion is very difficult to define. Listening to a musical se... more The relationship between music and emotion is very difficult to define. Listening to a musical selection can cause different individuals to respond in different ways. Measuring physiological responses to music covers only one part of how an individual experiences music. Listening to a favorite sports team's song often evokes feelings of pride and enthusiasm while a rival team's fan might feel antagonism. This type of induced emotion is obviously learned behavior. For this reason, rather than being a nature vs nurture discourse, research on emotions elicited by music is really a combination of nature and nurture.
Transactions on Engineering Technologies, 2015
Ray tracing is a widely used technique in rendering realistic scenes in Computer Graphics. Its ma... more Ray tracing is a widely used technique in rendering realistic scenes in Computer Graphics. Its main drawback has been that it is time consuming, requiring the rendering to finish from hours to sometimes days. For decades, the goal has been to speed up the processing of these scenes. Two popular grid traversal techniques have emerged: (a) Three Dimensional Digital Differential Analyzer (3DDDA) and (b) the Proximity Cloud (PC), which is a variation of 3DDDA. Both of these techniques try to limit the number of collision tests, which can be the most time consuming part of the algorithm. While both techniques allow impressive speedups, large dynamical varying scenes topology still challenge the real time rendering process. These techniques are optimal on static scenes, but object movement forces recalculation of the scene. This is a problem when using CPUs because parallelization is not easily available. Running these on GPUs, however, allows for parallelization. Apart from briefly summarizing some of our previous results from Ryan and Semwal (Proceedings of the world congress on engineering and computer science 2014, San Francisco, pp. 376–381 [1]), we also look to answer some of the more relevant questions about ray tracing, and what future holds for this area.
2014 IEEE International Conference on Systems, Man, and Cybernetics (SMC), 2014
Although video games have become interesting entertainment tools for millions of people, gamers w... more Although video games have become interesting entertainment tools for millions of people, gamers with mobility impairment face difficulties to be able to interact with games successfully because of impairment conditions. Even though developers working in the area of disbility provide special games, tools, and devices support interacting with video games, there is a lack of sufficient tools and devices that are compatible with certain motor impairment conditions. EEG-headset based brain computer interface system is alternative suggested tool which can support mobility gamers by using their brain activities, facial muscles, and eye movement to interact with games. This paper aims to examine and investigate the NIA controller's capability, which use EEG-headset, with two methods providing game accessibility. The first method reduces some keyboard switches and uses the NIA controller, and the other method eliminates the entire keyboard switches and uses only the NIA controller. The result shows that playing games with using the NIA controller and one- keyboard switch is done in successful and precise way. In contrast, playing games with using only the NIA controller could be done successfully but less precisely. Our conclusion is that the NIA controller could be combined with other accessible tools to improve the accessibility of games.
2013 IEEE Digital Signal Processing and Signal Processing Education Meeting (DSP/SPE), 2013
In Virtual Reality (VR) applications, an important measure of quality is the extent to which the ... more In Virtual Reality (VR) applications, an important measure of quality is the extent to which the user believes he is a part of (or immersed in) the virtual world which is being simulated. In this paper our focus is sound that is emanating from a source positioned in a particular 3D location in the real world. The ability to determine the 3D position of a sound source is called sound localization. A key element in localization is the Head Related Transfer Function (HTRF) and the associated impulse response. Here we use the U.C. Davis HTRF database along with a real-time DSP engine to implement the localization filters. By detecting the pose of the listener's head, the apparent position of a sound source relative to the listener is adjusted to accommodate the movement of the listener's head, thereby making the sound source appear to be at a fixed position in 3D space.
2013 IEEE 14th International Conference on Information Reuse & Integration (IRI), 2013
This paper reviews the overall impact of culture on the design of crossover applications, which a... more This paper reviews the overall impact of culture on the design of crossover applications, which are particularly intended to support the blind and visually impaired (B/VI) community. We believe that cultural differences have an impact on the proliferation and wide usage of any assistive technology. Therefore, cultural aspects must be considered in the design of crossover applications for the B/VI community. Comprehensive cultural vision is necessary for an innovative application to be accepted widely. In addition, there is a need to revise and periodically review the application's acceptance, and update a product based on the cultural changes that may occur to some communities especially with the vast cultural exchange. This paper reviews and highlights good design practices for assistive technology from cultural perspectives.
Proceeding Visualization '91
Page 1. Volume Rendering of Flow-Visualization Point Data Paul Gene Swann Sudhanshu Kumar Semwal ... more Page 1. Volume Rendering of Flow-Visualization Point Data Paul Gene Swann Sudhanshu Kumar Semwal Department of Computer Science University of Colorado Colorado Springs, CO 80933-7150 Department of Computer ...
2013 IEEE 14th International Conference on Information Reuse & Integration (IRI), 2013
This paper proposes the XCP Mobile Manager (iXMM): a new framework for remote management of Xen C... more This paper proposes the XCP Mobile Manager (iXMM): a new framework for remote management of Xen Cloud Platform (XCP) via mobile smart phones as an example of wearable computing. Mobile smart phones provide an opportunity to easily manage cloud virtualization, creating new forms of interactions. iXMM enhances both the portability and the usability of XCP management over the existing management tools. New forms of interactions, such as touch interfaces, are expected to emerge. For example, system administrators can control XCP remotely using smart phones with minimum Internet connection (i.e. 2G mobile internet connection). The implementation installs iXMM client as a native mobile application (currently iOS application) and allows access to iXMM through mobile web browsers. The evaluation shows that iXMM is very efficient in terms of the required computational resources including processing, memory, and Internet connection bandwidth; this makes it suitable for mobile computing environments. Even though full functionalities have not been implemented, iXMM provides an opportunity to complete similar tasks using smart phones, a trend which is expected to benefit aged population with minimal mobility.
Uploads
Papers by Sudhanshu Semwal