Publications

DeepSmooth: Efficient and Smooth Depth Completion

Published in CVPR-W 2023: VOCVALC, 2023

Accurate and consistent depth maps are essential for numerous applications across domains such as robotics, Augmented Reality and others. High-quality depth maps that are spatially and temporally consistent enable tasks such as Spatial Mapping, Video Portrait effects and more generally, 3D Scene Understanding. Depth data acquired from sensors is often incomplete and contains holes whereas depth estimated from RGB images can be inaccurate. This work focuses on Depth Completion, the task of filling holes in depth data using color images. Most work in depth completion formulates the task at the frame level, individually filling each frame’s depth. This results in undesirable flickering artifacts when the RGB-D video stream is viewed as a whole and has detrimental effects on downstream tasks. We propose DeepSmooth, a model that spatio-temporally propagates information to fill in depth maps. Using an EfficientNet and pseudo 3D-Conv based architecture, and a loss function which enforces consistency across space and time, the proposed solution produces smooth depth maps.

Krishna, Sriram & Vandrotti, Basavaraja Shanthappa. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 3357–3366, 2023 https://openaccess.thecvf.com/content/CVPR2023W/VOCVALC/html/Krishna_DeepSmooth_Efficient_and_Smooth_Depth_Completion_CVPRW_2023_paper.html

Using Sampling to Estimate and Improve Performance of Automated Scoring Systems with Guarantees

Published in The 12th Symposium on Educational Advances in Artificial Intelligence (EAAI-AAAI22) (in proceedings), 2021

Automated Scoring (AS), the natural language processing task of scoring essays and speeches in an educational testing setting, is growing in popularity and being deployed across contexts from government examinations to companies providing language proficiency services. However, existing systems either forgo human raters entirely, thus harming the reliability of the test, or score every response by both human and machine thereby increasing costs. We target the spectrum of possible solutions in between, making use of both humans and machines to provide a higher quality test while keeping costs reasonable to democratize access to AS. In this work, we propose a combination of the existing paradigms, sampling responses to be scored by humans intelligently. We propose reward sampling and observe significant gains in accuracy (19.80% increase on average) and quadratic weighted kappa (QWK) (25.60% on average) with a relatively small human budget (30% samples) using our proposed sampling. The accuracy increase observed using standard random and importance sampling baselines are 8.6% and 12.2% respectively. Furthermore, we demonstrate the system’s model agnostic nature by measuring its performance on a variety of models currently deployed in an AS setting as well as pseudo models. Finally, we propose an algorithm to estimate the accuracy/QWK with statistical guarantees (Our code is available at this https URL).

Singla, Y. K., Krishna, S., Shah, R. R., & Chen, C. (2021). Using Sampling to Estimate and Improve Performance of Automated Scoring Systems with Guarantees. arXiv preprint arXiv:2111.08906. https://arxiv.org/abs/2111.08906

Searching a Raw Video Database using Natural Language Queries

Published in International Conference on Advances in Electrical, Computing, Communication and Sustainable Technologies (ICAECT), 2021

The number of videos being produced and consequently stored in databases for video streaming platforms has been increasing exponentially over time. This vast database should be easily indexable to find the requisite clip or video to match the given search specification, preferably in the formof a textual query. This work aims to provide an end-to-end pipeline to search a video database with a voice query from theend user. The pipeline makes use of Recurrent Neural Networks in combination with Convolutional Neural Networks to generate captions of the video clips present in the database.

Krishna, Sriram, Vinay, Siddarth & Katharguppe, Srinivas. ICAECT 2021. https://arxiv.org/pdf/2012.15565

Gestop: Customizable Gesture Control of Computer Systems

Published in 8TH ACM IKDD CODS AND 26TH COMAD, 2021

The established way of interfacing with most computer systems is a mouse and keyboard. Hand gestures are an intuitive and effective touchless way to interact with computer systems. However, hand gesture based systems have seen low adoption among end-users primarily due to numerous technical hurdles in detecting in-air gestures accurately. This paper presents Gestop, a framework de- veloped to bridge this gap. The framework learns to detect gestures from demonstrations, is customizable by end-users and enables users to interact in real-time with computers having only RGB cameras, using gestures.

Krishna, Sriram & Sinha, Nishant. 8TH ACM IKDD CODS AND 26TH COMAD. 2021. 405-409 https://arxiv.org/pdf/2010.13197

Genetic Bi-objective Optimization Approach to Habitability Score

Published in MMLA 2019, 2019

The search for life outside the Solar System is an endeavour of astronomers all around the world. With hundreds of exoplanets be ing discovered due to advances in astronomy, there is a need to classify the habitability of these exoplanets. This is typically done using various metrics such as the Earth Similarity Index or the Planetary Habitability Index. In this paper, Genetic Algorithms are used to evaluate the best possible habitability scores using the Cobb-Douglas Habitability Score. Genetic Algorithm is a classic evolutionary algorithm used for solving op timization problems. The working of the algorithm is established through comparison with various benchmark functions and its functionality is ex tended to Multi-Objective optimization. The Cobb-Douglas Habitability Function is formulated as a bi-objective as well as a single objective optimization problem to find the optimal values to maximize the Cobb- Douglas Habitability Score for a set of promising exoplanets.

Krishna, Sriram & Pentapati, Niharika. MMLA 2019. COMMUNICATIONS IN COMPUTER AND INFORMATION SCIENCE, VOL 1290. SPRINGER https://arxiv.org/pdf/2010.05494

Sriram Krishna