Robust and Efficient Post-Processing for Video Object Detection (REPP)

1 Oct 2020  ·  Alberto Sabater, Luis Montesano, Ana C. Murillo ·

Object recognition in video is an important task for plenty of applications, including autonomous driving perception, surveillance tasks, wearable devices or IoT networks. Object recognition using video data is more challenging than using still images due to blur, occlusions or rare object poses. Specific video detectors with high computational cost or standard image detectors together with a fast post-processing algorithm achieve the current state-of-the-art. This work introduces a novel post-processing pipeline that overcomes some of the limitations of previous post-processing methods by introducing a learning-based similarity evaluation between detections across frames. Our method improves the results of state-of-the-art specific video detectors, specially regarding fast moving objects, and presents low resource requirements. And applied to efficient still image detectors, such as YOLO, provides comparable results to much more computationally intensive detectors.

PDF Abstract

Datasets


  Add Datasets introduced or used in this paper

Results from the Paper


Task Dataset Model Metric Name Metric Value Global Rank Benchmark
Video Object Detection ImageNet VID YOLOv3 MAP 68.6 # 15
Video Object Detection ImageNet VID REPP + YOLOv3 MAP 75.1 # 14
Video Object Detection ImageNet VID REPP + FGFA MAP 80.1 # 11
Video Object Detection ImageNet VID REPP + SELSA (ResNet-101) MAP 84.2 # 5

Methods


No methods listed for this paper. Add relevant methods here