People

Arash Vahdat

Arash Vahdat is a senior manager, leading the generative AI research team at NVIDIA Research. Before joining NVIDIA, he was a research scientist at D-Wave Systems, working on generative learning and its applications in label-efficient training. Before D-Wave, Arash was a research faculty member at Simon Fraser University (SFU), where he led deep learning-based video analysis research and taught master courses on machine learning for big data. Arash’s current research areas include generative learning, representation learning, and efficient deep learning. A complete publication list is also available on Arash's personal website.

Research Area(s)
Artificial Intelligence and Machine Learning
Computer Vision
Generative AI
Main Field of Interest
Artificial Intelligence and Machine Learning
Google Scholar
https://scholar.google.ca/citations?user=p9-nlRIAAAAJ&hl=en

Publications

2023

Open-Vocabulary Panoptic Segmentation with Text-to-Image Diffusion Models
Jiarui Xu, Sifei Liu, Arash Vahdat, Wonmin Byeon, Xiaolong Wang, Shalini De Mello
IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2023
Hightlight top 10%

2022

LANA: Latency Aware Network Acceleration
Pavlo Molchanov, Jimmy Hall, Hongxu Danny Yin, Jan Kautz, Nicolo Fusi, Arash Vahdat
European Conference on Computer Vision (ECCV), 2022
Diffusion Models for Adversarial Purification
Weili Nie, Brandon Guo, Yujia Huang, Chaowei Xiao, Arash Vahdat, Anima Anandkumar
International Conference on Machine Learning (ICML), 2022
A-ViT: Adaptive Tokens for Efficient Vision Transformer
Hongxu Danny Yin, Arash Vahdat, Jose M. Alvarez, Arun Mallya, Jan Kautz, Pavlo Molchanov
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2022 (Ora…
Tackling the Generative Learning Trilemma with Denoising Diffusion GANs
Zhisheng Xiao, Karsten Kreis, Arash Vahdat
International Conference on Learning Representations (ICLR) 2022 (Spotlight)
Score-Based Generative Modeling with Critically-Damped Langevin Diffusion
Tim Dockhorn, Arash Vahdat, Karsten Kreis
International Conference on Learning Representations (ICLR), 2022 (Spotlight)

2021

Don’t Generate Me: Training Differentially Private Generative Models with Sinkhorn Divergence
Tianshi Cao, Alex Bie, Arash Vahdat, Sanja Fidler, Karsten Kreis
Neural Information Processing Systems (NeurIPS) 2021
Score-based Generative Modeling in Latent Space
Arash Vahdat, Karsten Kreis, Jan Kautz
Neural Information Processing Systems (NeurIPS) 2021
A Contrastive Learning Approach for Training Variational Autoencoder Priors
Jyoti Aneja, Alexander Schwing, Jan Kautz, Arash Vahdat
Neural Information Processing Systems (NeurIPS) 2021
Controllable and Compositional Generation with Latent-Space Energy-Based Models
Weili Nie, Arash Vahdat, Anima Anandkumar
Neural Information Processing Systems (NeurIPS) 2021
See through Gradients: Image Batch Recovery via GradInversion
Hongxu Danny Yin, Arun Mallya, Arash Vahdat, Jose M. Alvarez, Jan Kautz, Pavlo Molchanov
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2021
VAEBM: A Symbiosis between Variational Autoencoders and Energy-based Models
Zhisheng Xiao, Karsten Kreis, Jan Kautz, Arash Vahdat
International Conference on Learning Representations (ICLR), 2021 (Spotlight)

2020

UNAS: Differentiable Architecture Search Meets Reinforcement Learning
Arash Vahdat, Arun Mallya, Ming-Yu Liu, Jan Kautz
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2021
NVAE: A Deep Hierarchical Variational Autoencoder
Arash Vahdat, Jan Kautz
Neural Information Processing Systems (NeurIPS) 2020 (Spotlight)
Contrastive Learning for Weakly Supervised Phrase Grounding
Tanmay Gupta, Arash Vahdat, Gal Chechik, Xiaodong Yang, Jan Kautz, Derek Hoiem
European Conference on Computer Vision (ECCV) 2020 (Spotlight)
On the Distance between Two Neural Networks and the Stability of Learning
Jeremy Bernstein, Arash Vahdat, Yisong Yue, Ming-Yu Liu
Neural Information Processing Systems (NeurIPS) 2020