ybyangjing Goto Github PK

faster_rcnn_sunrgbd

faster r-cnn for sun-rgbd

fastphotostyle

Style transfer, deep learning, feature transform

fce

fcn

free connect your private network from anywhere

featuretransfer

fedweit

This is an official Tensorflow-2 implementation of Federated Continual Learning with Inter-Client Weighted Transfer

flownet3d

FlowNet3D: Learning Scene Flow in 3D Point Clouds

fm-bench

An Evaluation of Feature Matchers for Fundamental Matrix Estimation (BMVC 2019)

frustum-pointnets

Frustum PointNets for 3D Object Detection from RGB-D Data

handong1587.github.io

hmd

Detailed Human Shape Estimation from a Single Image by Hierarchical Mesh Deformation (CVPR2019 Oral)

how-to-ask-questions-the-smart-way

本文原文由知名Hacker Eric S. Raymond 所撰寫，教你如何正確的提出技術問題並獲得你滿意的答案。

iclr2019-openreviewdata

Script that crawls meta data from ICLR OpenReview webpage. Tutorials on installing and using Selenium and ChromeDriver on Ubuntu.

ijcai-2022-zla

Codes for IJCAI'2022 Paper: Zero-Shot Logit Adjustment

imgaug

Image augmentation for machine learning experiments.

inception-v3-to-detection-privacy

Recent research has shown that the ubiquitous use of cameras and voice monitoring equipment in a home environment raises privacy concerns and affects human mental health, even causing mental disorders. This condition is a major obstacle to the deployment of smart home systems for elderly or disabled care. The current study uses a social robot for privacy situation detection. When a privacy situation is detected, the robot turns the camera away from the users and stores the abstract information in a text file. we proposed a method of detection of privacy situations based on Convolution Neural Network for Social Robot (DPS-SR). Firstly, this paper designed a questionnaire aiming at the privacy protection and social robot, and then analysed the focused factors and its attention degree of people that related to the privacy information caused by using social robot. Then designed a algorithm to detect the privacy situations based on Convolution Neural Network, and then the hardware platform of social robot and the overall workflow were detailed, and then implemented the complete system of DPS-SR. If the robot system detects a privacy situation, it will turn around and store the abstracted information to text file. Considering the results of the questionnaire survey, six classes of home situations were designed, and the training data sets for deep feature extraction about those situations were collected, which includes 2580 pictures. In order to check the performance of proposed system, we designed three experiment and four different testing datasets, which involves 960 pictures. The testing results shows as follows:1) when the testing dataset and the training dataset share the same home environment, the situation recognition accuracy of the proposed system is distributed from 97.5% to 100%, which indicates that the change of people has little effect on the situation recognition accuracy. 2) When the testing dataset and the training dataset include the same people, the system put out the recognition accuracy between 85% and 95%. 3) When the people and the home environment are changed between the testing dataset and training dataset, the system’s situation recognition accuracy is distributed in 80%~97.5%. The proposed system report an average recognition accuracy of 94.69%, which indicates the systems has reasonable robustness.

indoorunderstanding_3dgp

The code implement the method described in the 3DGP paper published in CVPR13 (see README for full title).

inloc_demo

istio-definitive-guide

Istio:权威指南

keras-grad-cam

An implementation of Grad-CAM with keras

keras-vis

Neural network visualization toolkit for keras

kl-loss

Bounding Box Regression with Uncertainty for Accurate Object Detection (CVPR'19)

labelfusion

LabelFusion: A Pipeline for Generating Ground Truth Labels for Real RGBD Data of Cluttered Scenes

layout2im

Official PyTorch Implementation of Image Generation from Layout - CVPR 2019-1

lcrl-

Logically-Constrained Reinforcement Learning

leeml-notes

李宏毅《机器学习》笔记，在线阅读地址：https://datawhalechina.github.io/leeml-notes

lstm-cf

LSTM-CF: Unifying Context Modeling and Fusion with LSTMs for RGB-D Scene Labeling

maplab

An open visual-inertial mapping framework.

mask_rcnn-for-sun-rgb-d

Mask R-CNN for image segmentation of SUN RGB-D and NYU datasets

mixup

Implementation of the mixup training method