-
Megvii
- Nanjing, Jiangsu, China
Stars
Inpaint anything using Segment Anything and inpainting models.
YOLOX is a high-performance anchor-free YOLO, exceeding yolov3~v5 with MegEngine, ONNX, TensorRT, ncnn, and OpenVINO supported. Documentation: https://yolox.readthedocs.io/
[ECCV2022] Contrastive Vision-Language Pre-training with Limited Resources
A simple image search engine using CLIP feature.
An open source implementation of CLIP.
Open source deep learning based unsupervised image retrieval toolbox built on PyTorch🔥
The official implementation of [CVPR2022] Decoupled Knowledge Distillation https://arxiv.org/abs/2203.08679 and [ICCV2023] DOT: A Distillation-Oriented Trainer https://openaccess.thecvf.com/content…
ClearML - Auto-Magical CI/CD to streamline your AI workload. Experiment Management, Data Management, Pipeline, Orchestration, Scheduling & Serving in one MLOps/LLMOps solution
Repo for counting stars and contributing. Press F to pay respect to glorious developers.
PyTorch implementation of Multi-Label Image Recognition with Graph Convolutional Networks, CVPR 2019.
LIBLINEAR -- A Library for Large Linear Classification
repo that holds code for improving on dropout using Stochastic Delta Rule
PyTorch 0.4 implementation of Mask-RCNN that runs on linux/mac/win10 with GPU/CPU
Kaggle New York City Taxi Trip Duration Solution 17th solution