Abstract:Human detection is one of the long-standing computer vision tasks, and it has been a cornerstone for many real-world applications, such as photo album organization, video surveillance, and autonomous driving. Benefiting from deep learning technologies, such as convolutional neural networks and modern object detectors, have been achieving much improved accuracy in generic object detection tasks. In this paper, we aim to improve deep learning-based human detection. Our main idea is to exploit semantic context information for human detection by using deep-learnt semantic features provided by semantic segmentation masks. Segmentation masks play as an attention mechanism and enforce the detectors to focus on the image regions where potential object candidates are likely to appear. Meanwhile, the extra segmentation mask channel can also guide the convolutional kernels to automatically learn more discriminative features which make it easier to distinguish the background and foreground. We implement our methods with two popular detection frameworks, i.e., faster R-CNN and SSD and experimentally analyze the effectiveness of the proposed methods. Evaluation results on the widely used MS-COCO dataset and the very recent CrowdHuman dataset are provided. Our proposed methods outperform the baseline detectors and achieve better performance on highly occluded human detection.

A Two-Stage Human Body Detector on Depth Data

A new depth descriptor for pedestrian detection in RGB-D images

Multi-scale Human Detection Based on Window Gradient Potential Energy with Partial Occlusion Handling

Human face detection under weak light based on depth information

Video Based Head Detection and Tracking Surveillance System

Detecting More Objects Indoors: a Depth-guided Detector for Mobile Robots

Detector-in-Detector: Multi-Level Analysis for Human-Parts

Robust Real-Time Human Perception with Depth Camera.

Human Detection Method Based on Multi-Part Detector and Multi-Instance Learning

Fast Human Detection in RGB-D Images Based on Color-Depth Joint Feature Learning.

An HOG-CT Human Detector with Histogram-Based Search.

Human detection using depth information

Robust 3-D Human Detection in Complex Environments With a Depth Camera.

A Two-stage Detection Algorithm for Workwear Compliance in Power Construction Scenarios Based on Feature Guidance

Human Detection Aided by Deeply Learned Semantic Masks

Deep Person Detection in 2D Range Data

Physical Blob Detector and Multi-Channel Color Shape Descriptor for Human Detection.

An ultra-fast human detection method for color-depth camera

Asymmetric Adaptive Fusion in a Two-Stream Network for RGB-D Human Detection

Human-object Interaction Detection with Depth-Augmented Clues

Human Detection Based on Fusion of Histograms of Oriented Gradients and Main Partial Features