deformable-detr

detrex is a research platform for DETR-based object detection, segmentation, pose estimation and other visual recognition tasks.

[CVPR 2022] Official Pytorch code for OW-DETR: Open-world Detection Transformer

[AAAI 2024] Official code for "Hyp-OW: Exploiting Hierarchical Structure Learning with Hyperbolic Distance Enhances Open World Object Detection"

Detectron2 wrapper for Deformable DETR

Chemistry-informed deep learning (1D Deformable-DETR) for automated ¹H NMR multiplet detection: δ, coupling J, proton count and line width in one forward pass.

AI-powered maritime surveillance system that detects ships and oil spills from SAR satellite imagery in real time. Built with TransUNet, Deformable DETR, and a gigapixel image viewer for human-in-the-loop validation.

Enable clear WebRTC video calls with fast room joins, screen sharing, chat, reactions, and mobile-friendly layouts for smooth collaboration.

A Computer Vision Model Development toolkit. cvmd uses NumPy arrays as both input and output, aiming to provide a unified and concise model inference interface.

Fast Triton kernels for multi-scale deformable attention (MSDA) — the core operator behind Deformable DETR, DINO, and Mask2Former — plus a from-scratch course and hands-on labs

Fork — nvcr.io container with pinned +cu111 wheels so the deformable-attention ops compile, plus Slurm scripts

This project was carried out during an internship at FESB (Faculty of Electrical Engineering, Mechanical Engineering and Naval Architecture), University of Split. The goal was to develop an Artificial Intelligence model for olive detection on olive trees, comparing two state-of-the-art computer vision approaches:

Document layout analysis pipeline (bachelor thesis): Deformable DETR + Table Transformer + OCR to segment scanned pages into text, tables and figures