You can not select more than 25 topics Topics must start with a chinese character,a letter or number, can include dashes ('-') and can be up to 35 characters long.
 
 
 
 
deng 380c73b3e1 更新 'README.md' 1 year ago
ascend310_infer fisrt commit 2 years ago
scripts fisrt commit 2 years ago
src fisrt commit 2 years ago
README.md 更新 'README.md' 1 year ago
__init__.py fisrt commit 2 years ago
eval.py fisrt commit 2 years ago
export.py fisrt commit 2 years ago
postprocess.py fisrt commit 2 years ago
preprocess.py fisrt commit 2 years ago
requirements.txt fisrt commit 2 years ago
train.py fisrt commit 2 years ago

语义分割是自主车辆理解周围场景的关键技术。对于实际的自主车辆,不希望花费大量的推理时间来获得高精度的分割结果。使用轻量级架构(编码器解码器或双通道)或对低分辨率图像进行推理,最近的方法实现了非常快速的场景解析,甚至可以在单个1080Ti GPU上以100 FPS以上的速度运行。然而,在这些实时方法和基于膨胀主干的模型之间仍然存在明显的性能差距。 为了解决这个问题,受HRNet的启发,作者提出了一种具有深度高分辨率表示能力的深度双分辨率网络,用于高分辨率图像的实时语义分割,特别是道路行驶图像。作者提出了

Python C++ Shell Text

Contributors (2)