NumPy, TensorFlow and PyTorch implementation of human body SMPL model and infant body SMIL model.
Official implementation of CVPR2020 paper "VIBE: Video Inference for Human Body Pose and Shape Estimation"
SCAN: Learning to Classify Images without Labels (ECCV 2020)
:zap: Yolo universal target detection model combined with EfficientNet-lite, the calculation amount is only 230Mflops(0.23Bflops), and the model size is 1.3MB