Abstract

Finding or tracking the location of an object accurately is a crucial problem in defense applications, robotics and computer vision. Radars fall into the spectrum of high-end defense sensors or systems upon which the security and surveillance of the entire world depends. There has been a lot of focus on the topic of Multi Sensor Tracking in recent years, with radars as the sensors. The Indian Air Force uses a Multi Sensor Tracking (MST) system to detect flights pan India, developed and supported by BEL (Bharat Electronics Limited), a defense agency we are working with. In this paper, we describe our Machine Learning approach, which is built on top of the existing system, the Air force uses. For purposes of this work, we trained our models on about 13 million anonymized real Multi Sensor tracking data points provided by radars performing tracking activity across the Indian air space. The approach has shown an increase in the accuracy of tracking by 5 percent from 91 to 96. The model and the corresponding code were transitioned to BEL, which has been tested in their simulation environment with a plan to take forward for ground testing. Our approach comprises of 3 steps:(a) We train a Neural Network model and a CatBoost model and ensemble them using a Logistic Regression model to predict one type of error, namely Splitting error, which can help to improve the accuracy of tracking.(b) We again train a Neural Network model and a CatBoost model and ensemble them using a different Logistic Regression model to predict the second type of error, namely Merging error, which can further improve the accuracy of tracking.(c) We use cosine …

Abstract

Recent focus on transcriptomic studies in food crops like rice, wheat and maize provide new opportunities to address issues related to agriculture and climate change. Re-analysis of such data available in public domain supplemented with annotations across molecular hierarchy can be of immense help to the plant research community, particularly co-expression networks representing transcriptionally coordinated genes that are often part of the same biological process. With this objective, we have developed NetREx, a Network-based Rice Expression Analysis Server, that hosts ranked co-expression networks of Oryza sativa using publicly available messenger RNA sequencing data across uniform experimental conditions. It provides a range of interactable data viewers and modules for analysing user-queried genes across different stress conditions (drought, flood, cold and osmosis) and hormonal treatments (abscisic and jasmonic acid) and tissues (root and shoot). Subnetworks of user-defined genes can be queried in pre-constructed tissue-specific networks, allowing users to view the fold change, module memberships, gene annotations and analysis of their neighbourhood genes and associated pathways. The web server also allows querying of orthologous genes from Arabidopsis, wheat, maize, barley and sorghum. Here, we demonstrate that NetREx can be used to identify novel candidate genes and tissue-specific interactions under stress conditions and can aid in the analysis and understanding of complex phenotypes linked to stress response in rice.

Abstract

Crowdfunding is an efficient method for raising funds for projects. When used for non-excludable public projects, the process is termed Civic Crowdfunding (CC) and is an active research area. Researchers have analyzed CC in game-theoretic settings assuming that agents are interested in (and contribute to) a single public project. Gen- eralizing the existing single project theory to determine agents’ equilibrium contributions for multiple projects is non-trivial – espe- cially with budget-constrained agents. This work hypothesizes that the agents can learn their equilibrium contributions with repeated participation in multi-project CC. We model CC as a game to vali- date the hypothesis and build an RL-based simulator: EqC-Learner. We first show that EqC-Learner learns a policy that mimics equi- librium contributions in a single project case for the existing CC mechanisms. To validate EqC-Learner for the multi-project case, we present certain theoretical results for the general multi-project case. Via extensive simulation-based experiments, we show that the learned contributions in EqC-Learner follow all the available theo- retical analysis. Thus, we believe that such an RL-based simulator can learn equilibrium contributions for the general multi-project CC mechanism

Abstract

Smart cities play a vital role in limiting the ill-effects of rapid urbanization on the environment without compromising on benefits such as improving infrastructure, standard of living, and productivity. However, the collection, storage, and sharing of data from the plethora of sensor networks in a typical smart city deployment warrants a well-defined data platform architecture. In this paper, we propose a multi layer architecture compliant with the oneM2M standards and the Indian Urban Data Exchange (IUDX) framework. The proposed architecture consists of Data Monitoring (DML), Data Storage (DSL), and Data Exchange (DEL) layers. The DML employs oneM2M as the middleware platform to achieve interoperability. The DSL uses a multi-tenant architecture with multiple logical databases, enabling efficient and reliable data management. The DEL utilizes standard data schemas and open APIs of IUDX to avoid data silos, and enables secure data sharing. Further, we present a proof-of-concept implementation of our architecture deployed in a university campus using OM2M, PostgreSQL, and Django. Finally, simulations mimicking real-time data insertion and retrieval showed that the DML can handle 600 concurrent users with an average latency under 100 milli seconds. The DSL improved the latency compared to a single database architecture and the DEL could handle 100 concurrent users with zero failed requests.

Abstract

Transformers, which were introduced for solving the task of machine translation, have expanded their utility in multiple domains. A recent application of transformers is in solving elementary mathematics problems. In this paper, we use a hybrid tokenization technique for encoding the mathematics and science problems and answers, which is used to train the transformer. We compare the performance of our tokenization with that of the char-to-char tokenzation in solving various types of mathematics and science problems. We discuss the accuracy, memory usage, and time to train the model with proposed tokenization. The proposed tokenization shows higher accuracy for some problems, and requires lesser memory compared to char-to-char tokenization. We propose an extended dataset of science and mathematics problems that consists of billions of samples in questionanswer format in raw text. Code and Dataset: https: //github.com/misterpawan/scimat2

Abstract

Locomotion in a virtual environment within a limited physical space is a complex activity. There exist established techniques to support limitless natural walking in virtual environments. These include Redirected walking, Dynamic path generation, and Walk-Inplace technique, etc. PragPal is one such limitless path generation technique that supports natural walking in virtual environments. It is a novel software-based non-haptic locomotion technique. In this paper, we detail the enhancements to the existing PragPal path generation technique that addresses underlying issues in the technique like (1) path collision at angular turns, (2) effective usage of the physical play area, and (3) the ability to set path-width during path turns.

Abstract

Eye disorders like myopia, hypermetropia, and color blindness affect millions of people around the world. Early detection can help the affected persons take preventive measures against further degradation. However, early detection requires regular testing and visits to trained healthcare professionals to judge the level of disorder. The customization of the test requirements can simplify the method of diagnosis. This paper presents an approach towards developing a virtual reality based low-cost head-mounted device (HMD). The standalone HMD can detect eye disorders like visual acuity, astigmatism, color vision deficiency, and age-related macular degeneration, limiting the involvement of healthcare professionals. We did a user study to test the accuracy of our prototype against various optical test scenarios. The study revealed that the prototype achieved an overall accuracy of 93%. We further aim to deploy the device in economically vulnerable communities with limited access to healthcare infrastructure.

Abstract

FPGAs are increasingly significant for deploying convolutional neural network (CNN) inference models because of performance demands and power constraints in embedded and data centre applications. Object detection and classification are essential tasks in computer vision. You Only Look Once (YOLO) is a very efficient algorithm for object detection and classification with its variant Yolov3-tiny specially designed for embedded applications. This paper presents the FPGA accelerator for multiple precisions (FIXED-8, FIXED-16, FLOAT32) of YoloV3-tiny. We use a homogenous systolic array architecture with a synchronized pipeline adder tree for convolution, allowing it to be scalable for multiple variants of Yolo with a change in host driver. We evaluated the design on Terasic DE5a-Net-DDR4. The Fixed point (FP-8, FP-16) implementations attain a throughput of 57 GOPs/s (> 23%) and 46.16 GOPs/s (> 340 %). We synthesized the first FLOAT32 imnlementation attaining 11.22 GFLOPs/s.

Abstract

This paper presents a low cost, easy to fabricate, thin-film sensor. We used three different methods to make the thin film using two different base (film) materials: carbon and graphite. We briefly summarize the method and outcome of multi-layer thin film fabrication using spray coating, spin-coating, and drop-casting. The quality of the thin films is compared, and electrical resistance is characterized. The fabricated thin film is used to measure humidity and is characterized against a commercial humidity sensor (DHT22). The experimental details and results of humidity measurement are discussed. We were able to achieve resistance range of 5k - 1.5k ohms for 40 - 80 % RH, respectively. Plans to further optimize the sensitivity of the sensor, in future is discussed. Index Terms—thin film sensors, spray coating, spin-coating, drop-casting, low-cost sensor fabrication, humidity sensor

Abstract

The possibility of a vectorlike top partner decaying to a new colorless weak-singlet scalar or pseudoscalar has attracted some attention in the literature. We investigate the production of a weaksinglet charge-2=3 T quark that can decay to a spinless boson (Φ) and a top quark at the LHC. Earlier [A. Bhardwaj et al., Roadmap to explore vectorlike quarks decaying to a new scalar or pseudoscalar, arXiv:2203.13753.], we showed that in a large part of the parameter space, the T → tΦ and the loopinduced Φ → gg decays become the dominant decay modes for these particles. Here, we investigate the discovery prospects of the T quark in this region through the above decays. In particular, we focus on the pp → TT → ðtΦÞðtΦÞ → ðtðggÞÞðtðggÞÞ channel. Separating this signal from the huge Standard Model background is a challenging task, forcing us to employ a multivariate machine-learning technique. We find that the above channel can be a discovery channel of the top partner in the large part of the parameter space where the above decay chain dominates. Our analysis is largely model independent, and hence our results would be useful in a broad class of new physics models.