PyTorch 预训练 CV 骨干网络集合

本仓库汇聚了 PyTorch 官方网站上所有机器视觉预训练模型的骨干网络集合。主要包括在 ImageNet1K 数据集上进行预训练的多种卷积神经网络（CNN）和 Vision Transformer 等视觉模型。整个集合被划分为 V1 和 V2 两个子集，涵盖了多个经典和先进的视觉模型版本。这些预训练骨干网络为用户提供了在图像识别、目标检测、图像分割等任务中进行迁移学习的强大基础。同时，它也为研究者和从业者在不同场景中灵活应用这些预训练模型提供了便捷的选择。

This repository collects the backbone collections of all computer vision pre-trained models from the official PyTorch website. It mainly includes various visual models such as convolutional neural networks (CNNs) and Vision Transformers pre-trained on the ImageNet1K dataset. The entire collection is divided into two subsets, V1 and V2, covering multiple classic and state-of-the-art visual model versions. These pre-trained backbones provide users with a powerful foundation for transfer learning in tasks such as image recognition, object detection, and image segmentation. Meanwhile, it also offers a convenient choice for researchers and practitioners to flexibly apply these pre-trained models in various scenarios.

搜集汇总

背景与挑战

背景概述

该数据集是一个PyTorch预训练计算机视觉骨干网络集合，包含在ImageNet1K数据集上预训练的卷积神经网络和Vision Transformer等模型，分为V1和V2两个版本。它为图像识别、目标检测和图像分割等任务提供了迁移学习的基础，方便研究者和从业者灵活应用。