Rank Sort Loss for Object Detection and Instance Segmentation

by   Kemal Oksuz, et al.

We propose Rank Sort (RS) Loss, as a ranking-based loss function to train deep object detection and instance segmentation methods (i.e. visual detectors). RS Loss supervises the classifier, a sub-network of these methods, to rank each positive above all negatives as well as to sort positives among themselves with respect to (wrt.) their continuous localisation qualities (e.g. Intersection-over-Union - IoU). To tackle the non-differentiable nature of ranking and sorting, we reformulate the incorporation of error-driven update with backpropagation as Identity Update, which enables us to model our novel sorting error among positives. With RS Loss, we significantly simplify training: (i) Thanks to our sorting objective, the positives are prioritized by the classifier without an additional auxiliary head (e.g. for centerness, IoU, mask-IoU), (ii) due to its ranking-based nature, RS Loss is robust to class imbalance, and thus, no sampling heuristic is required, and (iii) we address the multi-task nature of visual detectors using tuning-free task-balancing coefficients. Using RS Loss, we train seven diverse visual detectors only by tuning the learning rate, and show that it consistently outperforms baselines: e.g. our RS Loss improves (i) Faster R-CNN by   3 box AP and aLRP Loss (ranking-based baseline) by   2 box AP on COCO dataset, (ii) Mask R-CNN with repeat factor sampling (RFS) by 3.5 mask AP (  7 AP for rare classes) on LVIS dataset; and also outperforms all counterparts. Code available at https://github.com/kemaloksuz/RankSortLoss


page 5

page 12


A Ranking-based, Balanced Loss Function Unifying Classification and Localisation in Object Detection

We propose average Localization-Recall-Precision (aLRP), a unified, boun...

Mask-aware IoU for Anchor Assignment in Real-time Instance Segmentation

This paper presents Mask-aware Intersection-over-Union (maIoU) for assig...

AP-Loss for Accurate One-Stage Object Detection

One-stage object detectors are trained by optimizing classification-loss...

SCNet: Training Inference Sample Consistency for Instance Segmentation

Cascaded architectures have brought significant performance improvement ...

Correlation Loss: Enforcing Correlation between Classification and Localization

Object detectors are conventionally trained by a weighted sum of classif...

Robust Adversarial Perturbation on Deep Proposal-based Models

Adversarial noises are useful tools to probe the weakness of deep learni...

Seeing without Looking: Contextual Rescoring of Object Detections for AP Maximization

The majority of current object detectors lack context: class predictions...

Please sign up or login with your details

Forgot password? Click here to reset