SSSegmentation: An Open Source Supervised Semantic Segmentation Toolbox Based on PyTorch.

[CVPR 2024] Official Implementation of Collaborating Foundation models for Domain Generalized Semantic Segmentation

The official code repository for paper "Rethinking the Encoding of Satellite Image Time Series".

Separate mask2former from framework

Fine-tuning Mask2Former on the FoodSeg103 dataset for food segmentation, with deployment and interactive inference via Gradio.

AgriSegment — A multi-modal plant segmentation suite for agricultural research. It offers four FastAPI-powered tools—Hybrid (SegFormer + SAM), Interactive (real-time SAM), Semantic (batch SegFormer), and Panoptic (Mask2Former)—making AI-driven plant segmentation accessible to non-programmers.

This repository contains the example of fine-tuning panoptic segmentation model.

Assistive robot teleoperation system for blind and visually impaired users using ROS 2, haptic feedback (PS5 DualSense), shared autonomy, and real-time computer vision.

Open source project for waste detection developed by students of postgraduate course Artificial Intelligence with Deep Learning UPC

Windows GUI pipeline for turning 360° video into scale-calibrated 3DGS training assets: frame review, YOLO/SAM masks, Metashape/SphereSfM/COLMAP Rig, AprilTag scale calibration, and LichtFeld/Postshot/Brush exports.

Hugging Face Object Detection, Instance Segmentation, Semantic Segmentation, Panoptic Segmentation in PyTorch Fine-tuning.

Official repo for "Instance Segmentation in Autonomous Log Grasping Using EfficientViT-SAM MP-Former" paper

SimpleGVI: A fast and easy-to-use tool for Green View Index (GVI) calculation based on Mask2Former.

This repository provides a tutorial and code for reproducing the data and results presented in the following publication: Automated Mapping of Post-Storm Roof Damage Using Deep Learning and Aerial Imagery: A Case Study in the Caribbean (Kucharczyk, Nesbit, & Hugenholtz, 2025, Remote Sensing).

Mask2Former TensorFlow

A unified, feed-forward model integrating the MapAnything 3D reconstruction backbone with Mask2Former for multi-view consistent panoptic segmentation. This approach provides feature-rich and metric accurate semantic maps for indoor robot navigation while eliminating the need for posed inputs and costly test-time optimization.

BEACON improves transparent object instance segmentation with boundary supervision and content-adaptive query initialization on top of Mask2Former.

Google Colab playground for Mask2Former — supports Semantic, Panoptic & Instance segmentation across ADE20K, Cityscapes, COCO, and Mapillary Vistas with no-code model selection.

📏 Show screen dimensions in a simple overlay for quick debugging of CSS and layout issues, helping you build better web designs effortlessly.