Metadata-Version: 2.1
Name: hmdata
Version: 0.0.6
Summary: A package for human data
Home-page: https://github.com/YIN95/Human-Data-Utils
Author: Wenjie YIn
Author-email: yinwenjie159@hotmail.com
License: MIT license
Description: # Human-Data-Utils
        
        ## List of Data
        
        * [TotalCapture](#TotalCapture)
        
        ## TotalCapture
        
        The TotalCapture dataset is designed for 3D pose estimation from markerless multi-camera capture, It is the first dataset to have fully synchronised muli-view video, IMU and Vicon labelling for a large number of frames (∼1.9M), for many subjects, activities and viewpoints.
        
        Please check the information on this page: [TotalCapture Dataset](https://cvssp.org/data/totalcapture/)
        
        ### Data Preparation
        
        To request access to the TotalCapture Dataset, or for other queries please contact: a.gilbert@surrey.ac.uk.
        
        Download the TotalCapture Dataset and prepare in the following format:
        ```
        datafolder/
            |->S1/
                |->gyro_mag/
                    |-> acting1_Xsens_AuxFields.sensors
                    |-> acting2_Xsens_AuxFields.sensors
                    |-> ...
                |->imu/
                    |-> s2_acting1_calib_imu_bone.txt
                    |-> s2_acting1_calib_imu_ref.txt
                    |-> s2_acting1_Xsens.sensors
                    |-> ...
                |->video/
                    |-> acting1/
                        |-> TC_S2_acting1_cam1.mp4
                        |-> ...
                        |-> TC_S2_acting1_cam8.mp4
                    |-> acting2/
                    |-> .../
                    |-> walking3/
            |->.../
            |->S5/
        ```
        
        ### Data Preprocess
        
        Convert all videos to images:
        ```shell
        bash video2images.sh data_dir num_of_processes
        ```
        <!-- bash video2images.sh /media/ywj/Data/totalcapture/totalcapture 4 -->
        <!-- bash video2images.sh /mnt/md0/yinw/project/data/totalcapture 4 -->
Platform: UNKNOWN
Classifier: Programming Language :: Python :: 3
Classifier: License :: OSI Approved :: MIT License
Requires-Python: >=3.5
Description-Content-Type: text/markdown
