• Corpus ID: 234095672

Probabilistic Ranking-Aware Ensembles for Enhanced Object Detections

  title={Probabilistic Ranking-Aware Ensembles for Enhanced Object Detections},
  author={Mingyuan Mao and Baochang Zhang and David S. Doermann and Jie Guo and Shumin Han and Yuan Feng and Xiaodi Wang and Errui Ding},
Model ensembles are becoming one of the most effective approaches for improving object detection performance already optimized for a single detector. Conventional methods directly fuse bounding boxes but typically fail to consider proposal qualities when combining detectors. This leads to a new problem of confidence discrepancy for the detector ensembles. The confidence has little effect on single detectors but significantly affects detector ensembles. To address this issue, we propose a novel… 
1 Citations

Figures and Tables from this paper

NENet: Monocular Depth Estimation via Neural Ensembles

This work carefully investigates into a phenomenon that existing methods tend to exhibit asymmetric errors, and constructs a two-level ensemble scheme, NENet, to integrate multiple predictions from diverse base predictors, which forms a more reliable depth estimator, which substantially boosts the performance over base Predictors.



Bounding Box Regression With Uncertainty for Accurate Object Detection

A novel bounding box regression loss that greatly improves the localization accuracies of various architectures with nearly no additional computation and allows us to merge neighboring bounding boxes during non-maximum suppression (NMS), which further improves the globalization performance.

Acquisition of Localization Confidence for Accurate Object Detection

IoU-Net learning to predict the IoU between each detected bounding box and the matched ground-truth is proposed, which improves the NMS procedure by preserving accurately localized bounding boxes.

DR Loss: Improving Object Detection by Distributional Ranking

This work proposes a novel distributional ranking (DR) loss, which mitigates the intra-class imbalance issue in background candidates but also improves the efficiency for the ranking algorithm.

You Only Look Once: Unified, Real-Time Object Detection

Compared to state-of-the-art detection systems, YOLO makes more localization errors but is less likely to predict false positives on background, and outperforms other detection methods, including DPM and R-CNN, when generalizing from natural images to other domains like artwork.

Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation

This paper proposes a simple and scalable detection algorithm that improves mean average precision (mAP) by more than 30% relative to the previous best result on VOC 2012 -- achieving a mAP of 53.3%.

Soft-NMS — Improving Object Detection with One Line of Code

Soft-NMS is proposed, an algorithm which decays the detection scores of all other objects as a continuous function of their overlap with M and improves state-of-the-art in object detection from 39.8% to 40.9% with a single model.

RepPoints: Point Set Representation for Object Detection

It is shown that an anchor-free object detector based on RepPoints can be as effective as the state-of-the-art anchor-based detection methods, with 46.5 AP and 67.4 $AP_{50}$ on the COCO test-dev detection benchmark, using ResNet-101 model.

FoveaBox: Beyond Anchor-based Object Detector

FoveaBox is an accurate, flexible and completely anchor-free framework for object detection that shows great robustness and generalization ability to the changed distribution of bounding box shapes when challenged by the stretched testing images.

CenterNet: Object Detection with Keypoint Triplets

This paper designs two customized modules named cascade corner pooling and center pooling, which play the roles of enriching information collected by both top-left and bottom-right corners and providing more recognizable information at the central regions, respectively.

SSD: Single Shot MultiBox Detector

The approach, named SSD, discretizes the output space of bounding boxes into a set of default boxes over different aspect ratios and scales per feature map location, which makes SSD easy to train and straightforward to integrate into systems that require a detection component.