can be used for training and testing
can be used for training
can be used for training and validation
can be used for training, validation and testing
can be used for visual localization of scenes
can be used for validation and testing
provides the pixel-wise groundtruth of the crosswalk dataset
can be used for training, validation and testing to study the robustness of semantic segmentation across systems
Panoramic Annular Semantic Segmentation (PASS) dataset can be used for validation and testing
The pretrained model on NYUDv2 of ACNet
include RGB and polarization data for Semantic segmentation of traffic scene (for training and validation)
Research on Unconstrained Face Recognition and Application in Blind Assisting（Thesis、PPT）
Fast Banknote Identification with RGB-D Sensor for the Visually Impaired（Thesis、PPT、Code）
To enhance the resolution of 3D sound.
3D reconstruction and recognition based on RGB-D images.
Object recognition algorithm based on deep learning.
Indoor visual localization and navigation (including ppt).
An interactive system design of face recognition (including ppt).
A method to obtain geometry and material information of objects based on differential motion theory under uncontrolled condition.
Gesture Recognition
Online banknote identification based on neural network (including ppt and program files).
Short-range obstacle warning based on depth sensor (including ppt).
Visual human-computer interaction technique based on depth-sound mapping.
Pedestrian detection based on deep learning.
Visual assistance technology for the blind based on 3D terrain detection.
Human-computer interface system design in the visual assistance for the blind.
The pattern recognition technique in the visual assistance for the blind.