研究人员推出了 Lite Any Stereo V2 (LAS2)，这是一系列用于高效零样本立体匹配的超快模型。LAS2 通过采用针对实际延迟优化的纯二维成本聚合框架和三阶段训练策略，挑战了更快模型能力更弱的观念。这种方法能够实现更平滑的合成到真实迁移和更高的可靠性。LAS2 模型在高效立体方法中展现了最先进的准确性，同时实现了显著更低的推理时间，其中 LAS2-H 在 H200 和 Orin 硬件上分别比 Fast-FoundationStereo 快 1.8 倍和 2.7 倍。 AI
影响 这项研究可以实现立体匹配能力在资源受限设备上更高效的部署。
排序理由 该集群包含一篇详细介绍新模型架构和计算机视觉训练策略的研究论文。
在 Hugging Face Daily Papers 阅读 →
AI 生成摘要 · Google Gemini · 来自 4 个来源。 我们如何撰写摘要 →
Lite Any Stereo V2 (LAS2) presents an efficient stereo matching approach that achieves state-of-the-art accuracy with significantly reduced latency through optimized architecture and training strategies.
arXiv:2506.21091v2 Announce Type: replace Abstract: Stereo matching has become an increasingly important component of modern autonomous systems. Developing deep learning-based stereo matching models that deliver high accuracy while operating in real-time continues to be a major c…
arXiv:2606.24457v1 Announce Type: new Abstract: Recent advances in stereo matching have achieved remarkable accuracy, but often rely on large models, heavy computation, or additional foundation-model priors, making them difficult to deploy on resource-constrained platforms. In co…
Recent advances in stereo matching have achieved remarkable accuracy, but often rely on large models, heavy computation, or additional foundation-model priors, making them difficult to deploy on resource-constrained platforms. In contrast, efficient stereo models offer faster inf…