Alert button
Picture for Hardik Shah

Hardik Shah

Alert button

Enhancing End-to-End Multi-Task Dialogue Systems: A Study on Intrinsic Motivation Reinforcement Learning Algorithms for Improved Training and Adaptability

Jan 31, 2024
Navin Kamuni, Hardik Shah, Sathishkumar Chintala, Naveen Kunchakuri, Sujatha Alla Old Dominion

Viaarxiv icon

Text-to-Sticker: Style Tailoring Latent Diffusion Models for Human Expression

Nov 17, 2023
Animesh Sinha, Bo Sun, Anmol Kalia, Arantxa Casanova, Elliot Blanchard, David Yan, Winnie Zhang, Tony Nelli, Jiahui Chen, Hardik Shah, Licheng Yu, Mitesh Kumar Singh, Ankit Ramchandani, Maziar Sanjabi, Sonal Gupta, Amy Bearman, Dhruv Mahajan

Viaarxiv icon

EgoVLPv2: Egocentric Video-Language Pre-training with Fusion in the Backbone

Jul 11, 2023
Shraman Pramanick, Yale Song, Sayan Nag, Kevin Qinghong Lin, Hardik Shah, Mike Zheng Shou, Rama Chellappa, Pengchuan Zhang

Figure 1 for EgoVLPv2: Egocentric Video-Language Pre-training with Fusion in the Backbone
Figure 2 for EgoVLPv2: Egocentric Video-Language Pre-training with Fusion in the Backbone
Figure 3 for EgoVLPv2: Egocentric Video-Language Pre-training with Fusion in the Backbone
Figure 4 for EgoVLPv2: Egocentric Video-Language Pre-training with Fusion in the Backbone
Viaarxiv icon

End-to-End Neural Network Compression via $\frac{\ell_1}{\ell_2}$ Regularized Latency Surrogates

Jun 13, 2023
Anshul Nasery, Hardik Shah, Arun Sai Suggala, Prateek Jain

Figure 1 for End-to-End Neural Network Compression via $\frac{\ell_1}{\ell_2}$ Regularized Latency Surrogates
Figure 2 for End-to-End Neural Network Compression via $\frac{\ell_1}{\ell_2}$ Regularized Latency Surrogates
Figure 3 for End-to-End Neural Network Compression via $\frac{\ell_1}{\ell_2}$ Regularized Latency Surrogates
Figure 4 for End-to-End Neural Network Compression via $\frac{\ell_1}{\ell_2}$ Regularized Latency Surrogates
Viaarxiv icon

DIME-FM: DIstilling Multimodal and Efficient Foundation Models

Mar 31, 2023
Ximeng Sun, Pengchuan Zhang, Peizhao Zhang, Hardik Shah, Kate Saenko, Xide Xia

Figure 1 for DIME-FM: DIstilling Multimodal and Efficient Foundation Models
Figure 2 for DIME-FM: DIstilling Multimodal and Efficient Foundation Models
Figure 3 for DIME-FM: DIstilling Multimodal and Efficient Foundation Models
Figure 4 for DIME-FM: DIstilling Multimodal and Efficient Foundation Models
Viaarxiv icon

Tell Your Story: Task-Oriented Dialogs for Interactive Content Creation

Nov 08, 2022
Satwik Kottur, Seungwhan Moon, Aram H. Markosyan, Hardik Shah, Babak Damavandi, Alborz Geramifard

Figure 1 for Tell Your Story: Task-Oriented Dialogs for Interactive Content Creation
Figure 2 for Tell Your Story: Task-Oriented Dialogs for Interactive Content Creation
Figure 3 for Tell Your Story: Task-Oriented Dialogs for Interactive Content Creation
Figure 4 for Tell Your Story: Task-Oriented Dialogs for Interactive Content Creation
Viaarxiv icon

Fighting FIRe with FIRE: Assessing the Validity of Text-to-Video Retrieval Benchmarks

Oct 10, 2022
Pedro Rodriguez, Mahmoud Azab, Becka Silvert, Renato Sanchez, Linzy Labson, Hardik Shah, Seungwhan Moon

Figure 1 for Fighting FIRe with FIRE: Assessing the Validity of Text-to-Video Retrieval Benchmarks
Figure 2 for Fighting FIRe with FIRE: Assessing the Validity of Text-to-Video Retrieval Benchmarks
Figure 3 for Fighting FIRe with FIRE: Assessing the Validity of Text-to-Video Retrieval Benchmarks
Figure 4 for Fighting FIRe with FIRE: Assessing the Validity of Text-to-Video Retrieval Benchmarks
Viaarxiv icon

VoLTA: Vision-Language Transformer with Weakly-Supervised Local-Feature Alignment

Oct 09, 2022
Shraman Pramanick, Li Jing, Sayan Nag, Jiachen Zhu, Hardik Shah, Yann LeCun, Rama Chellappa

Figure 1 for VoLTA: Vision-Language Transformer with Weakly-Supervised Local-Feature Alignment
Figure 2 for VoLTA: Vision-Language Transformer with Weakly-Supervised Local-Feature Alignment
Figure 3 for VoLTA: Vision-Language Transformer with Weakly-Supervised Local-Feature Alignment
Figure 4 for VoLTA: Vision-Language Transformer with Weakly-Supervised Local-Feature Alignment
Viaarxiv icon