VIML-CVDL/Resource

Benchmarks, Datasets, Papers

Resource

Benchmarks, Datasets, Papers

Benchmarks

• Domain Adaptation for Object Detection

  (Code)Github

  (Paper)Domain Adaptive Faster R-CNN for Object Detection in the Wild

  (Faster-RCNN based, image-level and instance-level alignment)

Datasets

• Cityscape

  (Dataset)Cityscape

  (2D object detection, 2D semantic segmentation, real clean datasets+synthetic fog/rain dataset)

• Seeing Through Fog

  (Paper)Seeing Through Fog Without Seeing Fog: Deep Multimodal Sensor Fusion in Unseen Adverse Weather

  (Dataset)SeeingThroughFog Dataset

  (3D semantic segmentation, 3D object detection, 2D semantic segmentation, 2D object detection, real clean/fog/rain/snow/nighttime dataset, multi-sensor)