Intellindust AI Lab

Pioneering Vision AI for Edge Intelligence

About Us

Intellindust AI Lab is dedicated to advancing computer vision and artificial intelligence for edge devices. With expertise in imaging systems, efficient deep learning, and multimodal AI, our mission is to create high-performance, reliable, and efficient AI solutions for real-world applications.

Research Directions

Vision-Language Models for Edge Devices
Lightweight Real-Time Object Detection
Multimodal Sensor Fusion and Embodied Perception
RAW-Domain Video Enhancement and Detection
Trustworthy and Interpretable AI

We are actively looking for talented interns and researchers. Contact: shenxi@intellindust.com.

Awards & Honors

Open-set Recognition ECCV 2024

1st place in Open-set Recognition challenge

Computer Vision Workshop, ECCV 2024

Low-light Detection CVPR 2024

1st place in Low-light Object Detection challenge

Physics-Based Vision meets Deep Learning Workshop, CVPR 2024

Low-light Segmentation CVPR 2024

3rd place in Low-light Instance Segmentation challenge

Physics-Based Vision meets Deep Learning Workshop, CVPR 2024

Few-shot Object Detection CVPR 2025

Honorable Mention in Foundational Few-Shot Object Detection challenge

Foundational FSOD Workshop, CVPR 2025

NITRE Denoising and Demosaic CVPR 2025

2nd Place in Track 1 Joint Denoising and Demosaicing challenge

New Trends in Image Restoration and Enhancement Workshop, CVPR 2025

Soccer Net CVPR 2025

2nd Place in Ball Action Spotting challenge

CVsports Workshop, CVPR 2025

Selected Publications

DEIMv2 Thumbnail

[DEIMv2] Real-Time Object Detection Meets DINOv3

arXiv 2025

[Project] [PDF] [Code]

DEIM: DETR with Improved Matching for Fast Convergence

CVPR 2025

[Project] [PDF] [Slides] [Code]

SURE: SUrvey REcipes for building reliable and robust deep networks

CVPR 2024

[Project] [PDF] [Code]

Contact

Email: shenxi@intellindust.com

GitHub: Intellindust-AI-Lab