image-matting

Tiny client-side background remover with WebGPU, WASM and smart alpha-matte refinement.

Subject extraction and cutout for stills and video — classical, Apple Vision, or an injected matte provider behind one API.

UV-powered wrapper CLI and experiment workspace for gated BRIA Fibo-Edit-RMBG background removal

Clearcut: automatic local image background removal with BiRefNet, NVIDIA CUDA, and a private browser workspace.

ViTMatte is a state-of-the-art image matting model. It leverages plain Vision Transformers (ViTs) to accurately estimate the foreground object in images and videos. We see ViTMatte, its architecture, practical implementation steps, and its contributions to th

Local anime and illustration background removal with precision alpha editing for Windows

[ICLR 2025] MPMat: A 3D and Instance-Aware Human Matting and Editing Framework with Multiplane Representation

Offline, privacy-first desktop background remover. Tauri 2 + Rust + ONNX (RMBG-1.4) + Svelte 5. 100% on-device — no uploads, no telemetry.

ComfyUI nodes for CorridorKey — a neural green/blue screen keyer, with the community VRAM work bundled in: 2.98 GB peak instead of ~23 GB. Non-commercial (CC BY-NC-SA 4.0 + Corridor Digital terms).

Granular-computing (GrC) image matting for mammogram mass segmentation - applied and evaluated on MIAS and CBIS-DDSM.

Lasso a product in a photo → clean cutout on white or transparent, full resolution, 100% in your browser (BiRefNet on WebGPU via Transformers.js).

🐾 macOS desktop pet: turn your own pet's photos/videos into a living desktop companion via fully-offline AI matting; reacts to your Claude Code sessions

Predicting an alpha matte from an image and a trimap.

Bayesian Matting, Image Matting, Image Processing

VideoBGRemover — графическое приложение для автоматического удаления фона с видео с помощью нейросети. Поддерживает замену фона на зелёный экран (green screen) или генерацию альфа-маски (mask) для дальнейшей обработки.

Background removal as a small HTTP service: isnet through ONNX Runtime in Rust, Pillow-exact, 76 MB image, non-root

ONNX export/inference and Transformers.js tooling for MODNet real-time trimap-free portrait matting

A unified benchmark of four prompt-based matting models on plant images, with failure analysis and RGB-guided MAM refinement.

Julia implementation of the python package rembg

A High-Efficient Research Development Toolkit for Image Segmentation Based on Pytorch.