Abhishek Aich

Researcher at NEC Laboratories America

About

AI/CV Research Scientist at NEC Laboratories America, building next-generation vision–language and agentic systems for real-world perception. Specializing in agentic workflows, multimodal RAG, and vision–language systems, with an emphasis on bridging foundation models and deployable solutions. I have 4+ years of industry experience taking research from 0→1 into production-oriented systems, and a strong first-author (main-track) publication record in NeurIPS, CVPR, ICCV, and ICLR.

Ph.D. (2023) from UC Riverside (Vision and Learning Group), advised by Prof. Amit K. Roy-Chowdhury.

Agentic AI Vision-language models Open-vocab perception Multimodal RAG

Experience

NEC Labs America

Researcher

2023 - Present

NEC Labs America

Intern

2022

MERL

Intern

2021

UII America

Intern

2020

CV Google Scholar GitHub LinkedIn

Research opportunities: I’m happy to collaborate on projects involving (not limited to) agentic AI, vision-language models, open-vocab perception, and multimodal RAG. If you are interested, please send me an email.

Updates

Selected highlights from recent activity

Workshops & Service

News

Selected Publications

See Google Scholar for complete list.

iFinder: Structured Zero-Shot Vision-Based LLM Grounding for Dash-Cam Video Reasoning

Manyi Yao, Bingbing Zhuang, Sparsh Garg, Amit Roy-Chowdhury, Christian Shelton, Manmohan Chandraker, Abhishek Aich

NeurIPS 2025

Paper

Mapillary Vistas Validation for Fine-Grained Traffic Signs: A Benchmark Revealing Vision-Language Model Limitations

Sparsh Garg, Abhishek Aich

ICCV Workshop (4th DataCV Workshop & Challenge), 2025

Paper Code/Dataset

Progressive Token Length Scaling in Transformer Encoders for Efficient Universal Segmentation

Abhishek Aich, Yumin Suh, Samuel Schulter, Manmohan Chandraker

ICLR 2025

Paper Code

Image-Specific Adaptation of Transformer Encoders for Compute-Efficient Segmentation

Manyi Yao, Abhishek Aich, Yumin Suh, Amit K. Roy-Chowdhury, Christian R. Shelton, Manmohan Chandraker

WACV Workshop (5th Workshop on Image/Video/Audio Quality Assessment in Computer Vision, VLM and Diffusion Model), 2026

Paper Code

Efficient Controllable Multi-Task Architectures

Abhishek Aich, Samuel Schulter, Amit K. Roy-Chowdhury, Manmohan Chandraker, Yumin Suh

ICCV 2023

Paper

Cross-Domain Video Anomaly Detection without Target Domain Adaptation

Abhishek Aich, Kuan-Chuan Peng, Amit K. Roy-Chowdhury

WACV 2023

Paper

Leveraging Local Patch Differences in Multi-Object Scenes for Generative Adversarial Attacks

Abhishek Aich, Shasha Li, Chengyu Song, M. Salman Asif, Srikanth Krishnamurthy, Amit K. Roy-Chowdhury

WACV 2023

Paper

GAMA: Generative Adversarial Multi-Object Scene Attacks

Abhishek Aich*, Calvin-Khang Ta*, Akash Gupta, Chengyu Song, Srikanth Krishnamurthy, M. Salman Asif, Amit K. Roy-Chowdhury

NeurIPS 2022

Paper Code

Poisson2Sparse: Self-Supervised Poisson Denoising From a Single Image

Calvin-Khang Ta*, Abhishek Aich*, Akash Gupta*, Amit K. Roy-Chowdhury

MICCAI 2022

Paper Code

Adversarial Attacks on Black Box Video Classifiers: Leveraging the Power of Geometric Transformations

Shasha Li*, Abhishek Aich*, Shitong Zhu, M. Salman Asif, Chengyu Song, Amit K. Roy-Chowdhury, Srikanth Krishnamurthy

NeurIPS 2021

Paper Code

Spatio-Temporal Representation Factorization for Video-based Person Re-Identification

Abhishek Aich, Meng Zheng, Srikrishna Karanam, Terrence Chen, Amit K. Roy-Chowdhury, Ziyan Wu

ICCV 2021

Paper

ALANET: Adaptive Latent Attention Network for Joint Video Deblurring and Interpolation

Akash Gupta, Abhishek Aich, Amit K. Roy-Chowdhury

ACM MM 2020

Paper Code

Non-Adversarial Video Synthesis with Learned Priors

Abhishek Aich*, Akash Gupta*, Rameswar Panda, Rakib Hyder, M. Salman Asif, Amit K. Roy-Chowdhury

CVPR 2020

Paper Code

Technical Reports / Notes

Edit requests/additions/corrections are welcome!

Writeup
State Space Models: Nuts and Bolts
A practical, intuition-first walkthrough.
Writeup
Elastic Weight Consolidation (EWC): Nuts and Bolts
A concise reference + implementation notes.