Technical Program
Thursday, 13 October 2022
T1-1: Network Pruning
- 12:10 PM - 12:35 PM: Acceleration-aware, Retraining-free Evolutionary Pruning for Automated Fitment of Deep Learning Models on Edge Devices
Speakers: Jeet Dutta (TCS Research); Arijit Mukherjee (TCS Research)*; Swarnava Dey (TCS Research, Tata Consultancy Services Ltd.); Arpan Pal (Tata Consultancy Services)
- 12:35 PM - 01:00 PM: Accurate and Efficient Channel pruning via Orthogonal Matching Pursuit
Spreakers: Kiran Purohit (IIT Kharagpur)*; Anurag Parvathgari (IIT Kharagpur); Soumi Das (Indian Institute of Technology, Kharagpur); Sourangshu Bhattacharya (IIT Kharagpur)
Thursday, 13 October 2022
T1-2: Novel ML Applications
- 02:40 PM - 03:05 PM: A Hybrid Planning System for Smart Charging of Electric Fleets
Speaker: Kshitij Garg (TCS Research); Ajay Narayanan (TCS Research); Prasant Misra (TCS Research)*; Arunchandar Vasan (TCS Research); Vivek Bandhu (TCS); Debarupa Das (TCS)
- 03:05 PM - 03:30 PM: Ensembling Deep Learning And CIELAB Color Space Model for Fire Detection from UAV images
Speaker: Yash Jain (Indian Institute of Technology Roorkee)*; Vishu Saxena (Indian Institute of Technology Roorkee); Sparsh Mittal (IIT Roorkee)
- 03:30 PM - 03:55 PM: Efficient Graph based Recommender System with Weighted Averaging of Messages
Speaker: Faizan Ahemad (Amazon India)*
Thursday, 13 October 2022
D2-1:
- 11:30 AM - 11:40 AM: COSWARA: A Web Application for Data Collection and Real-time Screening of SARS-CoV-2 Infected Subjects (In-person)
Speakers: Debarpan Bhattacharya, Neeraj Sharma, Srikanth Raj Chetupalli, Sriram Ganapathy (Indian Institute of Science)
- 11:40 AM - 11:50 PM: vaAnI - an AI-powered voice-enabled search engine for the visually-challenged community (In-person)
Speaker: Shirsha Ray Chaudhuri, Prudhvi Nokku, Thomson Reuters
- 11:50 AM - 12:00 PM: Flexible AutoML: Accelerating AutoML adoption across Amazon (In-person)
Speakers: Abhishek Divekar, Gaurav Manchanda (Amazon)
- 12:00 PM - 12:10 PM: Automating NLP workflows using AutoML NLP (In-person)
Speakers: Shipra Jain, Chandra Sekhar Gupta Aravapalli (Microsoft R&D India Pvt Ltd)
- 12:10 PM - 12:20 PM: MLOps with enhanced performance control and observability (In-person)
Speakers: Indradumna Banerjee, Dinesh Mr Ghanta, Prateek Katageri, Girish Nautiyal, Pradeep Sanchana, Atin Modi (Oracle)
- 12:20 PM - 12:30 PM: We Built a Wall, for the Better (Running Online ML Experiments at Scale) (Virtual)
Speakers: Shahram Anver, Krithika Sundararajan (Gojek)
Thursday, 13 October 2022
D2-2:
- 2:10 PM - 2:20 PM: EEG Data Visualisation for Blinking Eyes(In-person)
Speaker: Darshil Shah (MINRO)
- 2:20 PM - 2:30 PM: Sahasraaksha: Video analytics for smart campus surveillance. (In-person)
Speaker: Sriram G (MINRO)
- 2:30 PM - 2:40 PM: FlAi MLOps (In-person)
Speaker: Alurkar, Ruju N (FlAI)
- 2:40 PM - 2:50 PM: Building Reusable AI Systems (In-person)
Speaker: Vikas Raturi (Intuit)
- 2:50 PM - 3:00 PM: Recommendation system for ML model deployment (Virtual)
Speakers: John Barboza, Mohan Gandhi (Amazon)
Friday, 14 October 2022
T2-1: Probabilistic Learning Systems
- 12:10 PM - 12:35 PM: RIDEN: uniform density histogram via reference instance discriminator with entropy maximization for distribution shift detection
Speakers: Kei Yonekawa (KDDI Research, Inc.)*; Kazuhiro Saito (KDDI Research, Inc.); Mori Kurokawa (KDDI Research, Inc.)
- 12:35 PM - 01:00 PM: DEC-aided SM-OFDM: A Spatial Modulation System with Deep Learning based Error Correction
Spreakers: Harsh Verma (IIIT Delhi)*; Vivek Bohara (IIIT Delhi); Anubha Gupta (IIIT Delhi)
Friday, 14 October 2022
T2-2: Real-time ML Systems
- 02:40 PM - 03:05 PM: Hetero-Rec: Optimal Deployment of Embeddings for High-Speed Recommendations
Speakers: Chinmay N Mahajan (TCS); Ashwin Krishnan (TCS Research)*; Manoj Nambiar (TCS); Rekha Singhal (TCS)
- 03:05 PM - 03:30 PM: System Design for an Integrated Lifelong Reinforcement Learning Agent For Real-Time Strategy Games
Speakers: Indranil Sur (SRI International )*; Zachary A Daniels (SRI International); Aswin Raghavan (SRI International); Jesse Hostetler (SRI International); Abrar Rahman (SRI International); Michael R Piacentino (SRI International); Ajay Divakaran (SRI, USA); Roberto Corizzo (American University); Kamil Faber (AGH University of Science and Technology); Nathalie Japkowicz (American University); Michael Baron (); James S Smith (Georgia Institute of Technology); Sahana Joshi (Georgia Tech); Zsolt Kira (Georgia Institute of Technology); Tyler L Hayes (RIT); Christopher Kanan (University of Rochester); Gianmarco J Gallardo (RIT)
Friday, 14 October 2022
D3-2:
- 11:30 AM - 11:50 AM: Invited Talk 1- Through the Looking Glass
Speaker: Amit Saha, Principal Engineer, Cisco Meraki, Bangalore
- 11:50 AM - 12:10 PM: Invited Talk 2- Temporal Knowledge Graph Representation and Question Answering
Speaker: Soumen Chakrabarti, Professor, CSE, IIT Bombay, India
- 12:10 PM - 1:00 PM: Doctoral Student Talks
- COSWARA: A Respiratory Acoustics Based Point-of-care-testing Tool for SARS-CoV-2 Infection
Speakers: Debarpan Bhattacharya*; Neeraj Sharma; Srikanth Raj Chetupalli; Sriram Ganapathy (Indian Institute of Science)
- Supervised approaches for Language and Speaker Recognition
Speaker: Shreyas Ramoji*; Sriram Ganapathy (Indian Institute of Science)
- Representation Learning for Graph Retrieval
Speaker: Indradyumna Roy (IIT Bombay)*
- Multiple Novel Intent Detection
Speaker: Ankan Mullick (IIT Kharagpur)*
- MPCLeague: Robust MPC Platform for Privacy-Preserving Machine Learning
Speaker: Ajith Suresh (Technical University of Darmstadt )*
- Tensor Networks Assisted Quantum Computing for Applications in Machine Learning and Quantum Chemistry
Speaker: Harshdeep Singh (Indian Institute of Technology, Kharagpur)*
- LearnDefend: Learning to Defend against Backdoor Attacks on Federated Learning
Speaker: Kiran Purohit (IIT Kharagpur)*
Friday, 14 October 2022
D3-3:
- 2:00 PM - 2:20 PM: Invited Talk 3- ML at Amazon: Learnings for Prospective Applied Scientists
Speaker: Vineet Chaoji, Sr. Applied Science Manager, Amazon
- 2:20 PM - 3:30 PM: Doctoral Student Talks
- TMCOSS: Thresholded Multi-Criteria Online Subset Selection for Data-Efficient Autonomous Driving
Speaker: Soumi Das (Indian Institute of Technology, Kharagpur)*
- MeFakE: Fake News Detection on Social Media using a PSO-based and Memory-efficient Deep Neural Network
Speaker: Dr. Rohit Kumar Kaliyar*; Dr. Anurag Goswami (Bennett University)
- Reconfigurable Deep Learning Architectures for Wireless Channel Estimation
Speaker: Syed Asrar ul haq (Indraprastha Institute of Information Technology Delhi)*
- Intelligent and Reconfigurable Wireless PHY
Speaker: Neelam Singh (IIITD)*
- Class Incremental Learning in Plants
Speaker: Yash Mittal*; Uttam Kumar (IIIT Bangalore)
- Data-Efficient Class-Incremental Learning
Speaker: Jayateja Kalla*; Soma Biswas (Indian Institute of Science)
- Analog Computing Systems and Hybrid-Architectures for Machine Learning at Edge
Speaker: Pratik Kumar (Indian Institute of Science (IISc))*
- Self-supervised learning for Speaker Diarization in the context of Conversational Speech
Speaker: Prachi Singh*; Sriram Ganapathy (Indian Institute of Science)
- Edge based split-control using ML algorithms in Bilateral Teleoperation of Robots: Design and Analysis
Speaker: Muneeb Ahmed*; Brejesh Lall Arzad A Kherani (IIT Delhi)
- 3:30 PM - 4:30 PM: Panel Discussion
Life after PhD; Data-centric vs Model-centric AI
Speakers: Amit Saha, Chiranjib Bhattacharyya, Suparna Bhattacharyya, Vineet Chaoji
Saturday, 15 October, 11:30 AM - 01:00 PM
T3-1: Hardware Systems for AI
- 12:10 PM - 12:35 PM: Performance improvement of reinforcement learning algorithms for online 3D bin packing using FPGA
Speaker: Ashwin Krishnan
- 12:35 PM - 01:00 PM: Performance Evaluation of gcForest inferencing on multi-core CPU and FPGA
Sharyu Mukhekar
Saturday, 15 October, 02:00 PM - 04:45 PM
T3-2: Software Systems for AI
- 02:40 PM - 03:05 PM: Automated Deep Learning Model Partitioning for Heterogeneous Edge Devices
Speaker: Arijit Mukherjee
- 03:05 PM - 03:30 PM: Health Assurance: AI Model Monitoring Platform
Speaker: Anirban Ghosh
- 03:30 PM - 03:55 PM: Efficient Vector Store System for Python using Shared Memory
Speaker: Dhruv Patel
- 03:55 PM - 04:20 PM: Address Location Correction System for Q-commerce
Speaker: Yashwanth Reddy
- 04:20 PM - 04:45 PM: How Provenance helps Quality Assurance Activities in AI/ML Systems
Speaker: KS Kim