– Ameya Prabhu – Seizing the means of prediction (robustly)!

Interests: Data-Centric ML Sanity Checks & Baselining Unlearning Continual Learning Binarization & Pruning Others

Data-Centric Machine Learning

Lifelong Benchmarks: Efficient Model Evaluation in an Era of Rapid Progress, DMLR-W, ICLR 2024
Ameya Prabhu*, Vishaal Udandarao*, Philip Torr, Matthias Bethge, Adel Bibi, Samuel Albanie
[PDF] [Code] [Data] [Twitter 🧵] [Talk] [Slides]

Corrective Machine Unlearning, DMLR-W, ICLR 2024
Shashwat Goel, *Ameya Prabhu*, Philip Torr, Ponnurangam Kumaraguru, Amartya Sanyal
[PDF] [Code] [Twitter 🧵] [Talk] [Slides]

Inverse Scaling: When Bigger Isn’t Better, TMLR 2023
Ian R. McKenzie, Alexander Lyzhov, Michael Martin Pieler, Alicia Parrish, Aaron Mueller, Ameya Prabhu, Euan McLean, Xudong Shen, Joe Cavanagh, Andrew George Gritsevskiy, Derik Kauffman, Aaron T. Kirtland, Zhengping Zhou, Yuhui Zhang, Sicong Huang, Daniel Wurgaft, Max Weiss, Alexis Ross, Gabriel Recchia, Alisa Liu, Jiacheng Liu, Tom Tseng, Tomasz Korbak, Najoung Kim, Samuel R. Bowman, Ethan Perez
[PDF] [Code] [Twitter 🧵] [Blogpost]

From Categories to Classifier: Name-Only Continual Learning by Exploring the Web, DMLR-W, ICLR 2024
Ameya Prabhu*, Hasan Abed Al Kader Hammoud*, Ser-Nam Lim, Bernard Ghanem, Philip Torr, Adel Bibi
[PDF] [Code] [Twitter 🧵] [Blogpost] [Talk] [Slides]

Sampling Bias in Deep Active Classification: An Empirical Study, EMNLP 2019
Ameya Prabhu*, Charles Dognin*, Maneesh Singh
[PDF] [Code] [Poster]

Sanity Checks for Machine Learning

RanDumb: A Simple Approach that Questions the Efficacy of Continual Representation Learning, Arxiv
Ameya Prabhu, Shiven Sinha, Ponnurangam Kumaraguru, Philip HS Torr, Ozan Sener, Puneet K Dokania
Technical Report, February 2024
[PDF] [Code] [Talk] [Slides]

No Cost Likelihood Manipulation at Test Time for Making Better Mistakes in Deep Networks, ICLR 2021
Shyamgopal Karthik, Ameya Prabhu, Puneet K. Dokania, Vineet Gandhi
[PDF] [Code] [Talk & Slides]

Simple Unsupervised Multi-Object Tracking, Arxiv
Shyamgopal Karthik, Ameya Prabhu, Vineet Gandhi
[PDF] [Code]

GDumb: A Simple Approach that Questions Our Progress in Continual Learning, ECCV 2020 (Oral)
Ameya Prabhu, Philip H.S. Torr, Puneet K. Dokania
[PDF] [Code] [Talk] [Teaser] [Slides] [Slides (V2)]

Sampling Bias in Deep Active Classification: An Empirical Study, EMNLP 2019
Ameya Prabhu*, Charles Dognin*, Maneesh Singh
In EMNLP 2019
[PDF] [Code] [Poster]

Machine Unlearning

Corrective Machine Unlearning, DMLR-W, ICLR 2024
Shashwat Goel, *Ameya Prabhu*, Philip Torr, Ponnurangam Kumaraguru, Amartya Sanyal
[PDF] [Code] [Twitter 🧵] [Talk] [Slides]

Towards adversarial evaluations for inexact machine unlearning, Arxiv
Shashwat Goel, *Ameya Prabhu*, Amartya Sanyal, Ser-Nam Lim, Philip Torr, Ponnurangam Kumaraguru
[PDF] [Code] [Talk] [Slides]

Continual Learning

RanDumb: A Simple Approach that Questions the Efficacy of Continual Representation Learning, Arxiv
Ameya Prabhu*, Shiven Sinha*, Ponnurangam Kumaraguru, Philip HS Torr, Ozan Sener, Puneet K Dokania
[PDF] [Code] [Talk] [Slides]

Rapid Adaptation in Online Continual Learning: Are We Evaluating It Right?, ICCV 2023
Hasan Abed Al Kader Hammoud*, Ameya Prabhu*, Ser-Nam Lim, Bernard Ghanem, Philip Torr, Adel Bibi
[PDF] [Code] [Talk] [Slides]

Computationally Budgeted Continual Learning: What Does Matter?, CVPR 2023
Ameya Prabhu*, Hasan Abed Al Kader Hammoud*, Puneet K. Dokania, Philip H.S. Torr, Ser-Nam Lim, Bernard Ghanem, Adel Bibi
[PDF] [Code] [Talk] [Slides]

Real-Time Evaluation in Online Continual Learning: A New Hope, CVPR 2023 (Oral)
Yasir Ghunaim*, Adel Bibi*, Kumail Alhamoud, Motasem Alfarra, Hasan Abed Al Kader Hammoud, Ameya Prabhu, Philip H.S. Torr, Bernard Ghanem
[PDF] [Code] [Talk] [Slides] [Poster]

Online Continual Learning Without the Storage Constraint, Arxiv
Ameya Prabhu, Zhipeng Cai, Puneet Dokania, Philip Torr, Vladlen Koltun, Ozan Sener
[PDF] [Code] [Talk] [Slides]

Uncertainty Estimation & Active Learning

CLActive: Episodic Memories for Rapid Active Learning, CoLLAs 2022
Sri Aurobindo Munagala, Sidhant Subramanian, Shyamgopal Karthik, Ameya Prabhu, Anoop Namboodiri
[PDF]

Sampling Bias in Deep Active Classification: An Empirical Study, EMNLP 2019
Ameya Prabhu*, Charles Dognin*, Maneesh Singh
[PDF] [Code] [Poster]

Making Better Mistakes

No Cost Likelihood Manipulation at Test Time for Making Better Mistakes in Deep Networks, ICLR 2021
Shyamgopal Karthik, Ameya Prabhu, Puneet K. Dokania, Vineet Gandhi
[PDF] [Code]

Multi-Object Tracking

Simple Unsupervised Multi-Object Tracking, Arxiv
Shyamgopal Karthik, Ameya Prabhu, Vineet Gandhi
Technical Report, June 2020. Latest Update: June 2020
[PDF] [Code]

Binarization & Pruning

STQ-Nets: Unifying Network Binarization and Structured Pruning, BMVC 2020
Aurobindo Munagala, Ameya Prabhu, Anoop Namboodiri
[PDF] [Talk & Slides]

Exploring Binarization and Pruning for Convolutional Neural Networks
Ameya Prabhu
Master’s Thesis, IIIT-H
[PDF] [Slides]

Deep Expander Networks: Efficient Deep Networks from Graph Theory, ECCV 2018 (Oral)
Ameya Prabhu*, Girish Varma*, Anoop Namboodiri
[PDF] [Code] [Talk] [Slides] [Poster]

Hybrid Binary Networks: Optimizing for Accuracy, Efficiency and Memory, WACV 2018 (Oral)
Ameya Prabhu, Vishal Batchu, Rohit Gajawada, Aurobindo Munagala, Anoop Namboodiri
[PDF] [Code] [Talk] [Slides] [Poster]

Distribution-Aware Binarization of Neural Networks for Sketch Recognition, WACV 2018 (Oral)
Ameya Prabhu, Vishal Batchu, Aurobindo Munagala, Rohit Gajawada, Anoop Namboodiri
[PDF] [Code] [Talk] [Slides] [Poster]

Others

Towards sub-word level compositions for sentiment analysis of hindi-english code mixed text, COLING 2016
Ameya Prabhu*, Aditya Joshi*, Manish Shrivastava, Vasudeva Varma
[PDF] [Poster] [Code]

Learning clustered sub-spaces for sketch-based image retrieval, ACPR15 (Oral)
Koustav Ghosal, Ameya Prabhu, Riddhiman Dasgupta, Anoop Namboodiri
[PDF] [Slides]