hkchengrex/Cutie

Health v2 · maintenance only

Adoption not part of the score

1095 stars · 113 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

Cutie is a video object segmentation framework with object-level memory reading, a follow-up to XMem offering better consistency, robustness, and speed. It includes Python code for scripted VOS inference plus a GUI tool for interactive video segmentation with permanent memory support.

Use cases

• segment objects in videos with a mask prompt
• track and segment objects across video frames
• interactively edit video object masks via a GUI
• run semi-supervised video object segmentation in a Python project
• improve segmentation robustness against distractors in videos

When to choose

• you need state-of-the-art semi-supervised video object segmentation
• you want a fast, robust alternative to XMem or DeAOT
• you need an interactive video segmentation tool with object add/remove controls

When to avoid

• you need image-only segmentation rather than video
• you need Windows support (only tested on Ubuntu)
• you lack a CUDA-capable GPU for inference

Facets

library · maturity active

computer-vision image-processing machine-learning deep-learning computer-vision deep-learning machine-learning python video-object-segmentation pytorch cvpr2024 interactive-segmentation memory-network gui-tool video linux gpu

Member repositories

For agents

markdown · JSON · MCP: product_card(name="hkchengrex/Cutie")