Semantic segmentation models with 500+ pretrained convolutional and transformer-based backbones.
[CAAI AIR'24] Bilateral Reference for High-Resolution Dichotomous Image Segmentation
The collection of pre-trained, state-of-the-art AI models for ailia SDK
Build, train, optimize, and run computer vision models locally, from raw images to live inference. Open source, optimized for Intel XPU (CPU-only and CUDA also supported).
High-performance Vision library in Python. Scale your research, not boilerplate.
Image augmentation for computer vision. AGPL-3.0-only or commercial licensing.
Cross-platform DICOM viewer for CT, MRI & PET/CT with MPR, 3D, fusion, editable segmentation and local AI auto-segmentation via TotalSegmentator.
LightMIS: Ultra-Lightweight Medical Image Segmentation Without a Stage-Wise Decoder