Timezone: »

On Model Calibration for Long-Tailed Object Detection and Instance Segmentation
Tai-Yu Pan · Cheng Zhang · Yandong Li · Hexiang Hu · Dong Xuan · Soravit Changpinyo · Boqing Gong · Wei-Lun Chao

Tue Dec 07 08:30 AM -- 10:00 AM (PST) @

Vanilla models for object detection and instance segmentation suffer from the heavy bias toward detecting frequent objects in the long-tailed setting. Existing methods address this issue mostly during training, e.g., by re-sampling or re-weighting. In this paper, we investigate a largely overlooked approach --- post-processing calibration of confidence scores. We propose NorCal, Normalized Calibration for long-tailed object detection and instance segmentation, a simple and straightforward recipe that reweighs the predicted scores of each class by its training sample size. We show that separately handling the background class and normalizing the scores over classes for each proposal are keys to achieving superior performance. On the LVIS dataset, NorCal can effectively improve nearly all the baseline models not only on rare classes but also on common and frequent classes. Finally, we conduct extensive analysis and ablation studies to offer insights into various modeling choices and mechanisms of our approach. Our code is publicly available at https://github.com/tydpan/NorCal.

Author Information

Tai-Yu Pan (The Ohio State University)
Cheng Zhang (The Ohio State Unversity)
Yandong Li (university of central florida)
Hexiang Hu (University of Southern California)
Dong Xuan (Ohio State University)
Soravit Changpinyo (University of Southern California (USC))
Boqing Gong (Tencent AI Lab)
Wei-Lun Chao (Ohio State University (OSU))

More from the Same Authors