Beyond mAP: Re-evaluating and Improving Performance in Instance Segmentation with Semantic Sorting and Contrastive Flow

  title={Beyond mAP: Re-evaluating and Improving Performance in Instance Segmentation with Semantic Sorting and Contrastive Flow},
  author={Rohit Jena and Lukas Zhornyak and Nehal Doiphode and Vivek P. Buch and James C. Gee and Jianbo Shi},
. Top-down instance segmentation methods improve mAP by hedging bets on low-confidence predictions to match a ground truth. Moreover, the query-key paradigm of top-down methods leads to the instance merging problem. An excessive number of duplicate predictions leads to the (over)counting error, and the independence of category and localization branches leads to the naming error. The de-facto mAP metric doesn’t capture these errors, as we show that a trivial dithering scheme can simultaneously… 



SOLO: Segmenting Objects by Locations

A new, embarrassingly simple approach to instance segmentation in images by introducing the notion of "instance categories", which assigns categories to each pixel within an instance according to the instance's location and size thus nicely converting instance mask segmentation into a classification-solvable problem.

SSAP: Single-Shot Instance Segmentation With Affinity Pyramid

This work proposes a single-shot proposal-free instance segmentation method that requires only one single pass for prediction, based on a pixel-pair affinity pyramid, which computes the probability that two pixels belong to the same instance in a hierarchical manner.

Semantic Instance Segmentation with a Discriminative Loss Function

This work proposes an approach of combining an off-the-shelf network with a principled loss function inspired by a metric learning objective that encourages a convolutional network to produce a representation of the image that can easily be clustered into instances with a simple post-processing step.

K-Net: Towards Unified Image Segmentation

Without bells and whistles, K-Net surpasses all previous published state-of-the-art single-model results of panoptic segmentation on MS COCO test-dev split and semantic segmentations on ADE20K val split with 55.2% PQ and 54.3% mIoU, respectively.

Instances as Queries

For the first time, it is demonstrated that a simple end-to-end query based framework can achieve the state-of-the-art performance in various instance-level recognition tasks.

End-to-End Object Detection with Transformers

This work presents a new method that views object detection as a direct set prediction problem, and demonstrates accuracy and run-time performance on par with the well-established and highly-optimized Faster RCNN baseline on the challenging COCO object detection dataset.

Real-time Instance Segmentation with Discriminative Orientation Maps

A real-time instance segmentation framework termed OrienMask, which combines the concise design for mask representation and its effective integration with the anchor-based object detector, and is qualified under real- time conditions while maintaining competitive accuracy.

Evaluating Large-Vocabulary Object Detectors: The Devil is in the Details

A simple fix is presented to the default AP implementation, ensuring that it is independent across categories as originally intended, and a pooled version of AP (APPool) is considered that rewards properly calibrated detectors by directly comparing crosscategory rankings.

The Devil is in Classification: A Simple Framework for Long-tail Instance Segmentation

This work systematically investigates performance drop of the state-of-the-art two-stage instance segmentation model Mask R-CNN on the recent long-tail LVIS dataset, and unveils that a major cause is the inaccurate classification of object proposals.

Fully Convolutional Instance-Aware Semantic Segmentation

The first fully convolutional end-to-end solution for instance-aware semantic segmentation task, which achieves state-of-the-art performance in both accuracy and efficiency, wins the COCO 2016 segmentation competition by a large margin.