Maschinelles Lernen · GitHub

Jag: Direct Box Prediction for Efficient Visual Grounding

A PyTorch implementation of NeRF (Neural Radiance Fields) that reproduces the results.

TorchKM is a GPU-accelerated PyTorch-based library for kernel machines including kernel SVM with a focus on fast and integrated train + tune workflows.

• scikit-learn SVM - Tuning means an outer GridSearchCV loop that refits from scratch for every setting — minutes to hours. - Folds tuning into the algorithm and reuses matrix work across folds and settings. Often 100×+ faster end-to-end.
• ThunderSVM - Fast SVM fitting, but you still bolt on your own CV loop, and it's SVM-only. - One call trains and tunes; also does DWD, logistic, and quantile regression.

PyTorch code for "VL-Adapter: Parameter-Efficient Transfer Learning for Vision-and-Language Tasks" (CVPR2022)

Fast, modular reference implementation and easy training of Semantic Segmentation algorithms in PyTorch.

• Modular Design: easily construct customized semantic segmentation models by combining different components.
• Distributed Training: >60% faster than the multi-thread parallel method(nn.DataParallel), we use the multi-processing parallel method.

A PyTorch Implementation of Neural IMage Assessment

🦓 AI-powered OpenStreetMap tool for importing zebra crossings

[ultralytics v8.3.75][yolov8/yolo11-pose][WIDER FACE]Upgrade YOLO5Face to YOLO8Face and YOLO11Face