#安卓#NanoDet-Plus⚡Super fast and lightweight anchor-free object detection model. 🔥Only 980 KB(int8) / 1.8MB (fp16) and run 97FPS on cellphone🔥
翻译 - ⚡超级快速，轻巧的无锚物体检测模型。 1.8仅1.8mb并在手机上运行97FPS🔥
PaddleClas 是一个为工业界和学术界所准备的图像识别任务工具集
🍅🍅🍅YOLOv5-Lite: Evolved from yolov5 and the size of model is only 900+kb (int8) and 1.7M (fp16). Reach 15 FPS on the Raspberry Pi 4B~
翻译 - shufflev2-yolov5：更轻、更快、更容易部署。由yolov5进化而来，模型大小只有1.7M（int8）和3.3M（fp16）。当输入大小为320×320时，它可以在树莓派4B上达到10+ FPS~
Research and Production Oriented Speaker Verification, Recognition and Diarization Toolkit
RM Operation can equivalently convert ResNet to VGG, which is better for pruning; and can help RepVGG perform better when the depth is large.
Object Classification Training Framework
A PaddlePaddle version image model zoo.
#计算机科学#Deep Neural Network Architectures with dlib
#计算机科学#The basic course of artificial intelligence of BJTU, 12 kinds of animal recognition based on paddlepaddle framework
RepVGG: Making VGG-style ConvNets Great Again
