一个让深度学习算法摆脱相机型号束缚的开源工具库
nvTorchCam: An Open-source Library for Camera-Agnostic Differentiable Geometric Vision
- 抽象投影与反投影操作，统一处理多种相机模型
- 支持端到端可微分计算，训练模型可跨相机类型直接迁移
- 适合自动驾驶、房产拍摄等多相机场景的开发者使用
我们提出nvTorchCam，一个基于Apache 2.0许可的开源库，旨在使深度学习算法具备相机无关性。该库抽象了投影与反投影等关键相机操作，使开发者仅需实现一次算法即可适配多种相机模型，包括针孔、鱼眼及360度等距投影全景图，广泛应用于汽车与房产拍摄场景。基于PyTorch构建，nvTorchCam支持全可微分计算、GPU加速与批量处理，显著提升效率。此外，针对某一相机类型训练的深度学习模型可直接迁移到其他类型，无需修改。本文介绍nvTorchCam的功能特性，并提供代码示例与图示说明其使用方法。源码与安装指南可在GitHub页面 https://github.com/NVlabs/nvTorchCam 获取。
原文摘要 · Abstract (English)
We introduce nvTorchCam, an open-source library under the Apache 2.0 license, designed to make deep learning algorithms camera model-independent. nvTorchCam abstracts critical camera operations such as projection and unprojection, allowing developers to implement algorithms once and apply them across diverse camera models--including pinhole, fisheye, and 360 equirectangular panoramas, which are commonly used in automotive and real estate capture applications. Built on PyTorch, nvTorchCam is fully differentiable and supports GPU acceleration and batching for efficient computation. Furthermore, deep learning models trained for one camera type can be directly transferred to other camera types without requiring additional modification. In this paper, we provide an overview of nvTorchCam, its functionality, and present various code examples and diagrams to demonstrate its usage. Source code and installation instructions can be found on the nvTorchCam GitHub page at https://github.com/NVlabs/nvTorchCam.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。