目标检测,语义分割,实例分割 近期SOTA

mac2025-10-15  6

准确率和fps兼具的一些最新方法

目标检测

CenterNet

CornerNet

FCOS

双目3D目标检测:

Stereo R-CNN  https://github.com/HKUST-Aerial-Robotics/Stereo-RCNN

3DSMFCNN   https://github.com/libornovax/master_thesis_code

目标检测模型库:

Detectron2

mmDetectron

SimpleDet

语义分割

BiSeNet   https://github.com/CoinCheung/BiSeNet,   

                https://github.com/ooooverflow/BiSeNet

LEDNet   https://github.com/xiaoyufenfei/LEDNet

实例分割

SOLO:

EmbedMask

TensorMask

RetinaMask

YOLACT++

YOLACT: https://github.com/dbolya/yolact

Mask R-CNN  https://github.com/facebookresearch/maskrcnn-benchmark

弱监督全景分割   

https://github.com/qizhuli/Weakly-Supervised-Panoptic-Segmentation

最新回复(0)