no code implementations • 18 Mar 2024 • Nishant Jain, Arun S. Suggala, Pradeep Shenoy
In particular, we show that using hard-to-classify instances in the validation set has both a theoretical connection to, and strong empirical evidence of generalization.
no code implementations • 28 Nov 2023 • Anuj Srivastava, Karm Patel, Pradeep Shenoy, Devarajan Sridharan
Here, we address a fundamental challenge with selective classification during automated diagnosis with domain-shifted medical images.
no code implementations • 28 Oct 2023 • Rishabh Tiwari, Durga Sivasubramanian, Anmol Mekala, Ganesh Ramakrishnan, Pradeep Shenoy
Deep networks tend to learn spurious feature-label correlations in real-world supervised learning tasks.
1 code implementation • 18 May 2023 • Nathan Beck, Suraj Kothawade, Pradeep Shenoy, Rishabh Iyer
However, learning unbiased models depends on building a dataset that is representative of a diverse range of realistic scenarios for a given task.
no code implementations • 30 Jan 2023 • Rishabh Tiwari, Pradeep Shenoy
Simplicity bias is the concerning tendency of deep networks to over-depend on simple, weakly predictive features, to the exclusion of stronger, more complex features.
no code implementations • CVPR 2023 • Aditay Tripathi, Rishubh Singh, Anirban Chakraborty, Pradeep Shenoy
We show that our augmentations significantly improve classification accuracy and robustness measures on a range of datasets and neural architectures.
1 code implementation • 14 Dec 2022 • Kushal Chauhan, Rishabh Tiwari, Jan Freyberg, Pradeep Shenoy, Krishnamurthy Dvijotham
Concept bottleneck models (CBMs) are interpretable neural networks that first predict labels for human-interpretable concepts relevant to the prediction task, and then predict the final label based on the concept label predictions.
no code implementations • 12 Dec 2022 • Nishant Jain, Karthikeyan Shanmugam, Pradeep Shenoy
Predictive uncertainty-a model's self awareness regarding its accuracy on an input-is key for both building robust models via training interventions and for test-time applications such as selective classification.
no code implementations • 12 Dec 2022 • Nishant Jain, Pradeep Shenoy
Second, we learn an auxiliary scorer model that recovers the appropriate mixture of timescales as a function of the instance itself.
no code implementations • 14 Nov 2022 • Aditay Tripathi, Rishubh Singh, Anirban Chakraborty, Pradeep Shenoy
We also obtain gains of up to 28% and 8. 5% on natural adversarial and out-of-distribution datasets like ImageNet-A (for ViT-B) and ImageNet-R (for ViT-S), respectively.
no code implementations • 7 Oct 2022 • Soumyabrata Pal, Prateek Varshney, Prateek Jain, Abhradeep Guha Thakurta, Gagan Madan, Gaurav Aggarwal, Pradeep Shenoy, Gaurav Srivastava
We then study the framework in the linear setting, where the problem reduces to that of estimating the sum of a rank-$r$ and a $k$-column sparse matrix using a small number of linear measurements.
1 code implementation • 29 Aug 2022 • Barath Mohan Umapathi, Kushal Chauhan, Pradeep Shenoy, Devarajan Sridharan
We also show that our solutions work well with other types of generative models (generative flows and variational autoencoders) and that their efficacy is governed by each model's reliance on local dependencies.
no code implementations • 12 Jun 2022 • Kushal Chauhan, Soumya Chatterjee, Akash Reddy, Balaraman Ravindran, Pradeep Shenoy
The options framework in Hierarchical Reinforcement Learning breaks down overall goals into a combination of options or simpler tasks and associated policies, allowing for abstraction in the action space.
1 code implementation • CVPR 2022 • Rishubh Singh, Pranav Gupta, Pradeep Shenoy, Ravikiran Sarvadevabhatla
Our framework involves independent dense prediction of object category and part attributes which increases scalability and reduces task complexity compared to the monolithic label space counterpart.
1 code implementation • 7 Feb 2022 • Durga Sivasubramanian, Ayush Maheshwari, Pradeep Shenoy, Prathosh AP, Ganesh Ramakrishnan
In several supervised learning scenarios, auxiliary losses are used in order to introduce additional information or constraints into the supervised learning objective.
no code implementations • CVPR 2022 • Rishabh Tiwari, KrishnaTeja Killamsetty, Rishabh Iyer, Pradeep Shenoy
To address this, replay-based CL approaches maintain and repeatedly retrain on a small buffer of data selected across encountered tasks.
1 code implementation • CVPR 2022 • Kushal Chauhan, Barath Mohan U, Pradeep Shenoy, Manish Gupta, Devarajan Sridharan
Likelihoods computed by deep generative models (DGMs) are a candidate metric for outlier detection with unlabeled data.
no code implementations • 6 Jun 2021 • Ananye Agarwal, Pradeep Shenoy, Mausam
A key limitation is that such neural-to-symbolic models can only be trained end-to-end for tasks where the output space is symbolic.
no code implementations • 9 Dec 2020 • Soumya Chatterjee, Pradeep Shenoy
In decision making tasks under uncertainty, humans display characteristic biases in seeking, integrating, and acting upon information relevant to the task.
no code implementations • NeurIPS 2012 • Pradeep Shenoy, Angela J. Yu
We postulate that this ``impatience'' to go is a strategic adjustment in response to the implicit asymmetry in the cost structure of GNG: the NoGo response requires waiting until the response deadline, while a Go response immediately terminates the current trial.
no code implementations • NeurIPS 2010 • Pradeep Shenoy, Angela J. Yu, Rajesh P. Rao
Intelligent agents are often faced with the need to choose actions with uncertain consequences, and to modify those actions according to ongoing sensory processing and changing task demands.