Shruti Agarwal

Research Scientist

San Jose

I’m a research scientist in AI for Content Authenticity team. My primary research interest lies in multimedia forensics, image analysis, machine learning, and computer vision.

Previously, I was a postdoc at CSAIL with Bill Freeman. Simultaneously, I was also working as a Lecturer for Computer Vision course in MIDS program at Berkeley School of Information. I did my PhD from CS department of UC Berkeley advised by Prof. Hany Farid.

Publications

On the Coexistence and Ensembling of Watermarks

Petrov, Aleksandar., Agarwal, Shruti., Torr, Phil., Bibi, Adel., Collomosse, John. (Dec. 3, 2025)

International Conference on Neural Information Processing Systems (NeurIPS)

TrustMark: Robust Watermarking and Watermark Removal for Arbitrary Resolution Images

Agarwal, Shruti., Bui, Tu., Collomosse, John. (Oct. 19, 2025)

International Conference on Computer Vision (ICCV 2025)

Your Text Encoder Can Be An Object-Level Watermarking Controller

Devulapally, Naresh., Huang, Mingzhen., Asnani, Vishal., Agarwal, Shruti., Lyu, Siwei., Lokhande, Vishnu. (Oct. 18, 2025)

ICCV 2025

CustomMark: Customization of Diffusion Models for Proactive Attribution

Asnani, Vishal., Collomosse, John., Liu, Xiaoming., Agarwal, Shruti. (Oct. 18, 2025)

ICCV APAI workshop 2025

Self-Improving Diffusion Models With Synthetic Data

Alemohammad, Sina., Humayun, Ahmed., Agarwal, Shruti., Collomosse, John., Baraniuk, Richard. (Feb. 1, 2025)

ICLR Workshop on Scaling Self-Improving Foundation Models (SSI-FM)

ProMark: Proactive Diffusion Watermarking for Causal Attribution

Asnani, Vishal., Collomosse, John., Bui, Tu., Liu, Xiaoming., Agarwal, Shruti. (Jun. 17, 2024)

CVPR 2024

EKILA: Synthetic Media Provenance and Attribution for Generative Art

Balan, Kar., Agarwal, Shruti., Jenni, Simon., Parsons, Andy., Gilbert, Andrew., Collomosse, John. (Jun. 21, 2023)