Search

Human-Machine Interaction (HMI) systems have gained huge interest in recent years, with reference expression comprehension being one of the main challenges. Traditionally human-machine interaction has been mostly limited to speech and visual modalities. However, to allow for more freedom in interaction, recent works have proposed the integration of additional modalities, such as gestures in HMI systems. We consider such an HMI system with pointing gestures and construct a table-top object picking scenario inside a simulated virtual reality (VR) environment to collect data. Previous works for such a task have used deep neural networks to classify the referred object, which lacks transparency. In this work, we propose an interpretable and compositional model, crucial to building robust HMI systems for real-world application, based on a neuro-symbolic approach to tackle this task. Finally we also show the generalizability of our model on unseen environments and report the results., Comment: Appeared in the 37th Annual Conference of the Japanese Society for Artificial Intelligence, 2023

The coordinate measuring machine(CMM) has been the benchmark of accuracy in measuring solid objects from nearly past 50 years or more. However with the advent of 3D scanning technology, the accuracy and the density of point cloud generated has taken over. In this project we not only compare the different algorithms that can be used in a 3D scanning software, but also create our own 3D scanner from off-the-shelf components like camera and projector. Our objective has been : 1. To develop a prototype for 3D scanner to achieve a system that performs at optimal accuracy over a wide typology of objects. 2. To minimise the cost using off-the-shelf components. 3. To reach very close to the accuracy of CMM.

Over the past ten years, many different approaches have been proposed for different aspects of the problem of resources management for long running, dynamic and diverse workloads such as processing query streams or distributed deep learning. Particularly for applications consisting of containerized microservices, researchers have attempted to address problems of dynamic selection of, for example: types and quantities of virtualized services (e.g., IaaS/VMs), vertical and horizontal scaling of different microservices, assigning microservices to VMs, task scheduling, or some combination thereof. In this context, we argue that frameworks like simulated annealing are highly suitable for online navigation of trade-offs between performance (SLO) and cost, particularly when the complex workloads and cloud-service offerings vary over time. Based on a macroscopic objective that combines both performance and cost terms, annealing facilitates light-weight and coherent policies of exploration and exploitation. In this paper, we first give some background on simulated annealing and then experimentally demonstrate its usefulness for different case studies, including service selection for both a single type of workload (e.g., distributed deep learning) and a mixture of workload types (exploring a partially categorical set of options), and container sizing for microservice benchmarks. We conclude with a discussion of how the basic annealing platform can be applied to other resource-management problems, hybridized with other methods, and accommodate user-specified rules of thumb.

As the machine learning and systems communities strive to achieve higher energy-efficiency through custom deep neural network (DNN) accelerators, varied precision or quantization levels, and model compression techniques, there is a need for design space exploration frameworks that incorporate quantization-aware processing elements into the accelerator design space while having accurate and fast power, performance, and area models. In this work, we present QUIDAM, a highly parameterized quantization-aware DNN accelerator and model co-exploration framework. Our framework can facilitate future research on design space exploration of DNN accelerators for various design choices such as bit precision, processing element type, scratchpad sizes of processing elements, global buffer size, number of total processing elements, and DNN configurations. Our results show that different bit precisions and processing element types lead to significant differences in terms of performance per area and energy. Specifically, our framework identifies a wide range of design points where performance per area and energy varies more than 5x and 35x, respectively. With the proposed framework, we show that lightweight processing elements achieve on par accuracy results and up to 5.7x more performance per area and energy improvement when compared to the best INT16 based implementation. Finally, due to the efficiency of the pre-characterized power, performance, and area models, QUIDAM can speed up the design exploration process by 3-4 orders of magnitude as it removes the need for expensive synthesis and characterization of each design., Comment: 25 pages, 12 figures. arXiv admin note: substantial text overlap with arXiv:2205.13045, arXiv:2205.08648

As the machine learning and systems communities strive to achieve higher energy-efficiency through custom deep neural network (DNN) accelerators, varied bit precision or quantization levels, there is a need for design space exploration frameworks that incorporate quantization-aware processing elements (PE) into the accelerator design space while having accurate and fast power, performance, and area models. In this work, we present QADAM, a highly parameterized quantization-aware power, performance, and area modeling framework for DNN accelerators. Our framework can facilitate future research on design space exploration and Pareto-efficiency of DNN accelerators for various design choices such as bit precision, PE type, scratchpad sizes of PEs, global buffer size, number of total PEs, and DNN configurations. Our results show that different bit precisions and PE types lead to significant differences in terms of performance per area and energy. Specifically, our framework identifies a wide range of design points where performance per area and energy varies more than 5x and 35x, respectively. We also show that the proposed lightweight processing elements (LightPEs) consistently achieve Pareto-optimal results in terms of accuracy and hardware-efficiency. With the proposed framework, we show that LightPEs achieve on par accuracy results and up to 5.7x more performance per area and energy improvement when compared to the best INT16 based design., Comment: Accepted paper at the Machine Learning for Computer Architecture and Systems (MLArchSys) Workshop in conjunction with ISCA 2021. This is an extended version of arXiv:2205.08648

As the machine learning and systems community strives to achieve higher energy-efficiency through custom DNN accelerators and model compression techniques, there is a need for a design space exploration framework that incorporates quantization-aware processing elements into the accelerator design space while having accurate and fast power, performance, and area models. In this work, we present QAPPA, a highly parameterized quantization-aware power, performance, and area modeling framework for DNN accelerators. Our framework can facilitate the future research on design space exploration of DNN accelerators for various design choices such as bit precision, processing element type, scratchpad sizes of processing elements, global buffer size, device bandwidth, number of total processing elements in the the design, and DNN workloads. Our results show that different bit precisions and processing element types lead to significant differences in terms of performance per area and energy. Specifically, our proposed lightweight processing elements achieve up to 4.9x more performance per area and energy improvement when compared to INT16 based implementation., Comment: Accepted paper at the On-Device Intelligence Workshop in conjunction with MLSys Conference 2021

The field of autonomous robotics is growing at a rapid rate. The trend to use increasingly more sensors in vehicles is driven both by legislation and consumer demands for higher safety and reliable service. Nowadays, robots are found everywhere, ranging from homes, hospitals to industries, and military operations. Autonomous robots are developed to be robust enough to work beside humans and to carry out jobs efficiently. Humans have a natural sense of understanding of the physical forces acting around them like gravity, sense of motion, etc. which are not taught explicitly but are developed naturally. However, this is not the case with robots. To make the robot fully autonomous and competent to work with humans, the robot must be able to perceive the situation and devise a plan for smooth operation, considering all the adversities that may occur while carrying out the tasks. In this thesis, we present an autonomous mobile robot platform that delivers the package within the VNIT campus without any human intercommunication. From an initial user-supplied geographic target location, the system plans an optimized path and autonomously navigates through it. The entire pipeline of an autonomous robot working in outdoor environments is explained in detail in this thesis., Comment: 56 pages, Bachelor Thesis

Multimodal IR, spanning text corpus, knowledge graph and images, called outside knowledge visual question answering (OKVQA), is of much recent interest. However, the popular data set has serious limitations. A surprisingly large fraction of queries do not assess the ability to integrate cross-modal information. Instead, some are independent of the image, some depend on speculation, some require OCR or are otherwise answerable from the image alone. To add to the above limitations, frequency-based guessing is very effective because of (unintended) widespread answer overlaps between the train and test folds. Overall, it is hard to determine when state-of-the-art systems exploit these weaknesses rather than really infer the answers, because they are opaque and their 'reasoning' process is uninterpretable. An equally important limitation is that the dataset is designed for the quantitative assessment only of the end-to-end answer retrieval task, with no provision for assessing the correct(semantic) interpretation of the input query. In response, we identify a key structural idiom in OKVQA ,viz., S3 (select, substitute and search), and build a new data set and challenge around it. Specifically, the questioner identifies an entity in the image and asks a question involving that entity which can be answered only by consulting a knowledge graph or corpus passage mentioning the entity. Our challenge consists of (i)OKVQAS3, a subset of OKVQA annotated based on the structural idiom and (ii)S3VQA, a new dataset built from scratch. We also present a neural but structurally transparent OKVQA system, S3, that explicitly addresses our challenge dataset, and outperforms recent competitive baselines., Comment: Accepted at SIGIR 2021

We propose a method for automatic calibration of a traffic surveillance camera with wide-angle lenses. Video footage of a few minutes is sufficient for the entire calibration process to take place. This method takes in the height of the camera from the ground plane as the only user input to overcome the scale ambiguity. The calibration is performed in two stages, 1. Intrinsic Calibration 2. Extrinsic Calibration. Intrinsic calibration is achieved by assuming an equidistant fisheye distortion and an ideal camera model. Extrinsic calibration is accomplished by estimating the two vanishing points, on the ground plane, from the motion of vehicles at perpendicular intersections. The first stage of intrinsic calibration is also valid for thermal cameras. Experiments have been conducted to demonstrate the effectiveness of this approach on visible as well as thermal cameras. Index Terms: fish-eye, calibration, thermal camera, intelligent transportation systems, vanishing points

Defect engineering has arisen as a promising approach to tune and optimise the adsorptive performance of metal-organic frameworks. However, the balance between enhanced adsorption and structural stability remains an open question. Here both CO2 adsorption capacity and mechanical stability are calculated for the zirconium-based UiO-66, which is subject to systematic variations of defect scenarios. Modulator-dependence, defect concentration and heterogeneity are explored in isolation. Mechanical stability is shown to be compromised at high pressures where uptake is enhanced with an increase in defect concentration. Nonetheless this reduction in stability is minimised for reo type defects and defects with trifluoroacetate substitution. Finally, heterogeneity and auxeticity may also play a role in overcoming the compromise between adsorption and stability.

In the following, we present example illustrative and experimental results comparing fair schedulers allocating resources from multiple servers to distributed application frameworks. Resources are allocated so that at least one resource is exhausted in every server. Schedulers considered include DRF (DRFH) and Best-Fit DRF (BF-DRF), TSF, and PS-DSF. We also consider server selection under Randomized Round Robin (RRR) and based on their residual (unreserved) resources. In the following, we consider cases with frameworks of equal priority and without server-preference constraints. We first give typical results of a illustrative numerical study and then give typical results of a study involving Spark workloads on Mesos which we have modified and open-sourced to prototype different schedulers.

We consider scaling laws for maximal energy efficiency of communicating a message to all the nodes in a wireless network, as the number of nodes in the network becomes large. Two cases of large wireless networks are studied -- dense random networks and constant density (extended) random networks. In addition, we also study finite size regular networks in order to understand how regularity in node placement affects energy consumption. We first establish an information-theoretic lower bound on the minimum energy per bit for multicasting in arbitrary wireless networks when the channel state information is not available at the transmitters. Upper bounds are obtained by constructing a simple flooding scheme that requires no information at the receivers about the channel states or the locations and identities of the nodes. The gap between the upper and lower bounds is only a constant factor for dense random networks and regular networks, and differs by a poly-logarithmic factor for extended random networks. Furthermore, we show that the proposed upper and lower bounds for random networks hold almost surely in the node locations as the number of nodes approaches infinity., Comment: 40 pages, 4 figures, submitted to IEEE Transactions on Information Theory