Closed Jokoe66 closed 4 years ago
You can look at the conversion scripts, for example convert_cityperson_to_coco.py to see. In our case, for CityPersons, we only prune based on height >= 50 and not on the visibility, for achieving an overall decent performance across all splits.
OK, but I wonder if the current data augmentation (particularly the random crop) strategy is suitable in such case. For example, the cropped patch may contain no visible part of a heavily occluded person, which will introduce persons with visibility 0 and cause classification ambiguity.
By the way, I find that the implemented random crop seems to be problematic. The cropped patch satisfies when any bbox has an IoU greater than the min_iou. But in the current implementation, the cropped patch satisfies when all bboxes have IoUs greater than min_iou. See this issue.
I mean number of severely occluded cases for e.g (vis <40 %) are close to ~10 % in cityPersons and this is randomly cropping so I am not sure how much an impact it would practically have. Moreover, as far as I remember we did see empirically a small gain by incorporating this augmentation. Perhaps you can give it a shot without this augmentation as well.
Regarding the potential bug, we actually over looked it. Support appreciated.
Yeah. The actual impact of these problems is probably small. Thanks.
The training sets of CityPersons and Caltech benchmarks currently used in Pedestron are the Reasonable subsets (h>=50 and vis>=0.65), right ? Do you have any plan to train the detectors on other training subsets? Since when training occlusion-handling detectors (like JL-TopS, PDOE, MGAN), the subset (h>=50 and vis >=0.3) is commonly used.