Abstract:In complex traffic scenes, the orientation and location of pedestrians are important criteria for judging their intentions. We note that pedestrians are characterized by variability in appearance and small differences among orientations (especially adjacent orientations), thus causing general object detection algorithms to perform poorly in extracting features. So, extracting more discriminative features is an effective way to solve this problem. To this end, we propose a novel framework to enhance feature extraction involving pedestrian orientation detection (orientation classification and location regression). The framework consists of two modules, multi-window Transformer parallel fusion feature pyramid (MTPF) and gated graph (GG). The MTPF module is used for multi-layer feature fusion, which improves the feature representation of the prediction map by extracting high-level semantic information from deep layers and recovering missing contextual information from shallow layers. Specifically, it is achieved by setting a sliding window on multiple prediction maps and fused by the Transformer. The region proposal is abstracted into a graph with six nodes in the GG module, where each node represents a body part. We utilize GG to learn the spatial dependencies among body parts and learn features by aggregating information from neighbors. Finally, pedestrian orientation classification and location regression are performed on a graph containing rich relationships among nodes. According to the survey, there are currently no methods and datasets that can be directly used for pedestrian orientation detection, so we manually annotate pedestrian orientations on three public datasets containing a large number of pedestrian samples, and compare the proposed method with the current state-of-the-art object detection methods by comparison, the results demonstrate the effectiveness of the proposed method.

Continuous Pedestrian Orientation Estimation Using Human Keypoints.

Pedestrian Crossing Action Recognition and Trajectory Prediction with 3D Human Keypoints

Pedestrian movement trajectory reappearance and crossing feature expression based on video processing

Modeling social interaction and intention for pedestrian trajectory prediction

Using Appearance to Predict Pedestrian Trajectories Through Disparity-Guided Attention and Convolutional LSTM

A Real-Time Predictive Pedestrian Collision Warning Service for Cooperative Intelligent Transportation Systems Using 3D Pose Estimation

Pedestrian Trajectory Prediction in Heterogeneous Traffic Using Pose Keypoints-Based Convolutional Encoder-Decoder Network

Pedestrian Motion State Estimation From 2D Pose

Integrating Orientation Cue With EOH-OLBP-Based Multilevel Features for Human Detection

Deep Multi-Task Networks For Occluded Pedestrian Pose Estimation

Unified Pedestrian Detection And Orientation Recognition Using Single-Frame

Multi-window Transformer Parallel Fusion Feature Pyramid Network for Pedestrian Orientation Detection

Pedestrian Detection Directing at the Region of Interest in Videos

Algorithm of Pedestrian Pose Recognition Based on Keypoint Detection.

Pedestrian Trajectory Prediction in Heterogeneous Traffic using Facial Keypoints- based Convolutional Encoder-decoder Network

Pedestrian Detection and Tracking Using HOG and Oriented-Lbp Features

On-Board Pedestrian Trajectory Prediction Using Behavioral Features

Pedestrian Intention Prediction Via Depth Augmented Scene Restoration

CF-LSTM: Cascaded Feature-Based Long Short-Term Networks for Predicting Pedestrian Trajectory

A Computer Graphics-Based Framework for 3D Pose Estimation of Pedestrians