Analytics Softwares

Behavior Analysis for Surveillance

Our Behavior Analysis system is an AI-driven solution that detects and interprets human behaviors in real time. Designed for drone-based surveillance, it efficiently monitors vast or hard-to-reach areas and identifies complex anomalies such as violence, loitering, etc . Powered by Miphi semiconductors chips, the system runs heavy behavior analysis models on lower-end GPUs such as the RTX A6000, achieving faster inference through an optimized compute–time trade-off. These chips enable real-time processing and threat detection without the need for extensive high-end hardware resources.

Building on this foundation, our system incorporates Open-World Video Anomaly Detection (VAD) to handle unseen or evolving behaviors, ensuring adaptability in dynamic environments. By integrating machine learning and vision-language models, it continuously learns from live feeds, generates real-time alerts, and minimizes false positives for reliable situational awareness.

Agentic Path Planning

Our project focuses on enhancing drone usability through intelligent path planning to address issues like signal loss during long-range flights. We are reviewing and improving existing path planning algorithms—traditionally designed for 2D grids—to function efficiently in 3D environments. The goal is to enable future integration with machine learning models for autonomous flight control. Additionally, we are optimizing these algorithms for deployment on edge devices to ensure real-time performance and scalability.

Controlling the Behaviour of Agents in a Multi-Agent Collaboration

The work currently focused is mostly related to a single agent, but we already have a roadmap for how we will be integrating it in a multi-agent environment as well. The foundation of this area is the work currently being done by us in the single-agent domain. Key research topics in this area include dynamic formation maintenance for swarms of drones and collaborative behaviour specialized to the task at hand.
The goal of this domain of research is to make drones that not only have specialized roles, but also how multiple such entities can work together to perform more generalized roles as well.

Video summarisation and Video Question Answering

To enhance the decision-making capabilities of personnel monitoring drone feeds—either live through mobile ground control stations or during post-processing of stored footage—it is crucial to develop models that can generate concise summaries and extract answers to natural language questions. Video summarization plays a vital role in efficiently processing large volumes of visual data, ensuring that only the most relevant information is retained. A particularly effective approach involves leveraging submodular functions, which provide a lightweight framework for selecting representative frames while maintaining diversity and coverage. Video Question Answering (Video QA) can be effectively handled by Vision-Language Models (VLMs), which can be further enhanced using subset selection. We are also working on adapting these models to improve drone-based surveillance and monitoring.

Publications

Selected publications related to the projects.

SeRVo-HOI: Skew-Robust Human-Object Interactions in Videos

Apoorva Agarwal, Rishabh Dabral, Arjun Jain, Ganesh Ramakrishnan

In Proceedings of The 11th IEEE Winter Conference on Applications of Computer Vision (WACV 2023).

AUTOMATA: Gradient Based Data Subset Selection for Compute-Efficient Hyper-parameter Tuning

KrishnaTeja Killamsetty, Guttu Sai Abhishek, Aakriti, Ganesh Ramakrishnan, Alexandre V. Evfimievski, Lucian Popa, Rishabh K. Iyer

Accepted paper at the Thirty-sixth Conference on Neural Information Processing Systems (Neurips 2022).

SPEAR : Semi-supervised Data Programming in Python

Guttu Sai Abhishek, Harshad Ingole, Parth Laturia, Vineeth Dorna, Ayush Maheshwari, Rishabh Iyer, Ganesh Ramakrishnan

Accepted paper at the 2022 Conference on Empirical Methods in Natural Language Processing, EMNLP 2022, Abu Dhabi (Demo paper).

PRISM: A Rich Class of Parameterized Submodular Information Measures for Guided Data Subset Selection

Suraj Kothawade, Vishal Kaushal, Ganesh Ramakrishnan, Jeff Bilmes, Rishabh Iyer

In Proceedings of The 36th AAAI Conference on Artificial Intelligence (AAAI 2022).

GLISTER: Generalization based Data Subset Selection for Efficient and Robust Learning [source code , video]

Krishnateja Killamsetty, Durga S, Ganesh Ramakrishnan, and Rishabh Iyer

In Proceedings of The 35th AAAI Conference on Artificial Intelligence (AAAI 2021).

Training Data Subset Selection for Regression With Controlled Generalization Error [source code ]

Durga Sivasubramanian, Rishabh Iyer, Ganesh Ramakrishnan, Abir De

Accepted paper at the 38th International Conference on Machine Learning (ICML 2021).

GRAD-MATCH: Gradient Matching based Data Subset Selection for EfficientDeep Model Training [source code , video]

Krishnateja Killamsetty, Durga Sivasubramanian, Ganesh Ramakrishnan, Abir De, Rishabh Iyer

Accepted paper at the 38th International Conference on Machine Learning (ICML 2021).

Semi-Supervised Data Programming with Subset Selection [source code , video]

Ayush Maheshwari, Oishik Chatterjee, Krishnateja Killamsetty, Ganesh Ramakrishnan, Rishabh Iyer

Accepted paper at the 59th Annual Meeting of the Association for Computational Linguistics (ACL 2021 Findings).

Wisdom of (Binned) Crowds: A Bayesian Stratification Paradigm for Crowd Counting [source code ]

Sravya Shivapuja, Mansi Khamkar, Divij Bajaj, Ganesh Ramakrishnan, Ravi Kiran Sarvadevabhatla

Accepted at The 29th ACM International Conference on Multimedia (ACMM 2021)

CATALIST: CAmera TrAnsformations for multi-LIngual Scene Text recognition [project page, slides]

Shivam Sood, Rohit Saluja, Ganesh Ramakrishnan and Parag Chaudhuri

Accepted at ICDAR 2021 Workshop on Camera-Based Document Analysis and Recognition (CBDAR 2021, 9TH EDITION)

Select, Substitute, Search: A New Benchmark for Knowledge-Augmented Visual Question Answering [data and source code]

Aman Jain, Mayank Kothyari, Vishwajeet Kumar, Preethi Jyothi, Ganesh Ramakrishnan, Soumen Chakrabarti

Accepted paper at The 44th International ACM Conference on Research and Development in Information Retrieval (SIGIR), Resource Track, 2021.

Exploration of Spatial and Temporal Modeling Alternatives for HOI [presentation, source code]

Rishabh Dabral, Srijon Sarkar, Sai Praneeth Reddy, Ganesh Ramakrishnan

In Proceedings of The 9th IEEE Winter Conference on Applications of Computer Vision (WACV 2021).

LIGHTEN: Learning Interactions with Graph and Heirarchical TEmporal Networks for HOI in videos [source code]

Sai Praneeth Sunkesula, Rishabh Dabral, Ganesh Ramakrishnan

In Proceedings of The 28th ACM International Conference on Multimedia (ACMM 2020), Seattle, USA.

Caption Alignment for Low Resource Audio-Visual Data

Vighnesh Reddy Konda, Mayur Warialani, Rakesh Prasanth Achari, Varad Bhatnagar, Japrakash Akula, Preethi Jyothi, Ganesh Ramakrishnan, Gholamreza Haffari and Pankaj Singh

Accepted paper at the 21st INTERSPEECH Conference (Interspeech 2020), Shanghai, China

A Framework towards Domain Specific Video Summarization

Vishal Kaushal, Sandeep Subramanian, Suraj Kothawade, Rishabh Iyer, Ganesh Ramakrishnan

Accepted paper at the 7th IEEE Winter Conference on Applications of Computer Vision (WACV), 2019, Hawaii, USA.

A Framework towards Domain Specific Video Summarization

Vishal Kaushal, Sandeep Subramanian, Suraj Kothawade, Rishabh Iyer, Ganesh Ramakrishnan

Accepted paper at the 7th IEEE Winter Conference on Applications of Computer Vision (WACV), 2019, Hawaii, USA.

Learning From Less Data: Diversified Subset Selection and Active Learning in Image Classification Tasks

Vishal Kaushal, Rishabh Iyer, Anurag Sahoo, Khoshrav Doctor, Narasimha Raju, Ganesh Ramakrishnan

Accepted paper at the 7th IEEE Winter Conference on Applications of Computer Vision (WACV), 2019, Hawaii, USA.

Demystifying Multi-Faceted Video Summarization: Tradeoff Between Diversity,Representation, Coverage and Importance

Vishal Kaushal, Rishabh Iyer, Anurag Sahoo, Pratik Dubal, Suraj Kothawade, Rohan Mahadev, Kunal Dargan, Ganesh Ramakrishnan

Accepted paper at the 7th IEEE Winter Conference on Applications of Computer Vision (WACV), 2019, Hawaii, USA.

Anomaly Detection in Surveillance Videos

Sukalyan Bhakat, Ganesh Ramakrishnan

Best demo paper award. In Proceedings of the ACM India Joint International Conference on Data Science and Management of Data, CoDS-COMAD '19, Kolkota, India.

An Interactive Multi-Label Consensus Labeling Model for Multiple Labeler Judgments

Ashish Kulkarni, Narasimha Raju Uppalapati, Pankaj Singh, Ganesh Ramakrishnan

In Proceedings of the 32nd AAAI Conference on Artificial Intelligence (AAAI-18), New Orleans, Louisiana, USA.

Synthesis of Programs from Multimodal Datasets

Shantanu Thakoor, Simoni Shah, Ganesh Ramakrishnan, Amitabha Sanyal

In Proceedings of the 32nd AAAI Conference on Artificial Intelligence (AAAI-18), New Orleans, Louisiana, USA.

Error Detection and Corrections in Indic OCR using LSTMs

Rohit Saluja, Devaraj Adiga, Parag Chaudhuri, Ganesh Ramakrishnan and Mark Carman

International Conference on Document Analysis and Recognition (ICDAR) 2017, Kyoto, Japan.

A Framework for Document Specific Error Detection and Corrections in Indic OCR

Rohit Saluja, Devaraj Adiga, Ganesh Ramakrishnan, Parag Chaudhuri and Mark Carman

1st International Workshop on Open Services and Tools for Document Analysis (ICDAR- OST) 2017, Kyoto, Japan.

Improving the learnability of classifiers for Sanskrit OCR corrections

Devaraja Adiga, Rohit Saluja, Vaibhav Agrawal, Ganesh Ramakrishnan, Parag Chaudhuri, K. Ramasubramanian and Malhar Kulkarni

Proceedings of the 17th World Sanskrit Conference, Vancouver, 2018.

A Framework for Error Detection and Corrections in Sanskrit

Rohit Saluja, Devaraj Adiga, Parag Chaudhuri, Ganesh Ramakrishnan and Mark Carman

Research and Innovation Symposium in Computing (RISC) 2017 (Most Admiring Poster Presentation Award), IIT-Bombay, India.

Summarizing Multi-Document Topic Hierarchies using Submodular Mixtures

Ramakrishna Bairi, Rishabh Iyer, Ganesh Ramakrishnan and Jeff Bilmes

In Proceedings of the Annual Meeting of the Association for Computational Linguistics (ACL), Beijing, China, July - 2015

Video Analytics for Defense and Security Applications

Data Efficient Machine Learning (DECILE)

Document Analytics

Interested in our work?

Technology on Video Survelliance Analytics licensed to SrivisifAI Technologies Pvt Ltd (Send email)

Video Analytics (won The National eGovernance Gold Award 2022)

Surakshavyuha

Video Surveillance Analytics

Jigyasa

Video Search Analytics

Videosummy

Visiocity Summarization Benchmark

Video Summarization Software

Drishti

Compliance and Quality Monitoring

Data Efficient Machine Learning (DECILE)

Document Analytics

IndicOCR

Optical Character Recognition for Indian Texts

Maharashtra Drone Mission (MDM)

Behavior Analysis for Surveillance

Agentic Path Planning

Controlling the Behaviour of Agents in a Multi-Agent Collaboration

Video summarisation and Video Question Answering

Publications

In The News

MiPhi Semiconductors and IIT Bombay Join Forces to Drive AI-Powered Drone Innovation

Mumbai, 28 July 2025

Team of Prof. Ganesh Ramakrishnan (with Dr. Vishal Kaushal and SrivisifAI Technologies) wins National e-Governance Award

25th National Conference on eGovernance, 26th-27th November, Jammu

Prof. Ganesh Ramakrishnan and Vishal Kaushal receive the Dr. P. K. Patwardhan Technology Development Award 2020

IIT Bombay, September 6, 2021

CCTV tech by IIT-Bombay checks footage, sends out alerts

Mid-Day, Mumbai, June 18, 2021

IIT-Bombay develops AI platform for real-time video surveillance

Hindustan Times, Mumbai, June 17, 2021

IIT-Bombay develops AI-based solutions for video analytics, surveillance

Times Of India, Mumbai, June 17, 2021

Director, IIT Bombay

Facebook, June 16, 2021

TEAM

Contact Us

Interested in our work?