Salman Asif

M. Salman Asif Associate Professor Department of Electrical and Computer Engineering Cooperating Faculty Department of Computer Science and Engineering University of California, Riverside Email: sasif@ucr.edu Phone: (951) 827-2385 Office: 4126 Multidisciplinary Research Building (MRB) Mail: 900 University Ave., Riverside, CA 92521 Web: www.ece.ucr.edu/~sasif Google Scholar profile

I am an Associate Professor in the Department of Electrical and Computer Engineering at the University of California, Riverside.

Research interests. My research interests broadly lie in the areas of computational imaging, signal/image processing, computer vision, and machine learning. I design algorithms to learn and exploit latent signal structures for fast and efficient information recovery.

Current research focus is on computational imaging (physics-based and data-driven approaches), adversarial attacks and defenses for robust computer vision, multi-task and multi-modal machine learning, and wearable sensors for health monitoring and rehabilitation.

News and updates

• March 2023: Two papers accepted to CVPR 2023. Congratulations to Zikui, Yaoteng, Qi, and Zhan!

  • Ensemble-based Blackbox Attacks on Dense Prediction
    

  • DNeRV: Modeling Inherent Dynamics via Difference Neural Representation for Videos

• February 2023: Two papers accepted in ICASSP 2023

  • Compressive Sensing with Tensorized Autoencoder
    

  • Privacy Preserving Face Recognition with Lensless Camera

• January 2023: Two journal papers on coded illuminations and lensless imaging. Congratulations to Yucheng!

  • Coded Illumination for Improved Lensless Imaging
    

  • Coded Illumination for 3D Lensless Imaging

• January 2023: One journal paper on resolution analysis of masks-based lensless cameras.

  • Spatial and Axial Resolution Limits for Mask-based Lensless Cameras

• September 2022: Two papers accepted to NeurIPS 2022. Congratulations to lead students Zikui, Abhishek, and all the collaborators!

  • Blackbox Attacks via Surrogate Ensemble Search
    TLDR: Optimizing a weighted loss function over surrogate ensemble provides highly successful and query efficient blackbox (targeted and untargeted) attacks.

  • GAMA: Generative Adversarial Multi-Object Scene Attacks
    TLDR: A joint vision/language model (such as CLIP) provides image and text embeddings. The generative model learns to generate perturbed images that are close to image embedding and far from the text embedding of the clean images.

• August 2022: One paper on Leveraging Local Patch Differences in Multi-Object Scenes for Generative Adversarial Attacks accepted to WACV 2023 (Round 1).

• July 2022: Our paper on Incremental Task Learning with Incremental Rank Updates accepted to ECCV 2022. Congratulations to Rakib, Ken, Boyu and the collaborators!

• June 2022: One paper on Physics-Guided Neural-Network-Based Inverse Design of a Photonic-Plasmonic Nanodevice for Superfocusing accepted to ACS Applied Materials & Interfaces (joint work with Ming Liu group).

• May 2022: One paper on Low-Light Color Imaging Via Cross-Camera Synthesisaccepted to IEEE JSTSP (joint work with Zhan Ma group).

• March 2022: One paper accepted to CVPR 2022. Congratulations to Zikui and all the collaborators!

  • Zero-Query Transfer Attacks on Context-Aware Object Detectors (joint work with Zikui Cai, Shantanu Rane, Alejandro Brito, Srikanth Krishnamurthy, Chengyu Song, and Amit Roy-Chowdhury)

• more news