Deeplabcut

**New features:**
- new human project demo: extract human pose from video with pretrained network (deeplabcut.create_pretrained_human_project). The weights are from the DeeperCut paper.
- new COLAB demo on human project
- new skeleton analyzer (and docs) see 349
- multi peak extraction see 321

**Function updates:**
- documentation updates
- updated COLAB notebooks
- updated testscripts
- tensorflow code clean up (warning messages for TF 1.13 and 1.14)

**Function updates:**
-- GUI update to include auto-off zoom/pan and lock view 293
-- GUI up/down hotkeys added for radio buttons 318 319
-- implemented robuster downloading of the weights (and legacy version intact) 280
-- new cropping functionality for plotting on full frame or cropped ``frame:deeplabcut.create_labeled_video(path_config_file,[video],displaycropped=True)`` 301 309
-- creating labeled video with cropping+save_frames=True 218
-- deeplabcut.filterpredictions includes a median filter (beyond ARIMA)
-- new feature to plot "skeleton" and a history of points on videos during ``deeplabcut.create_labeled_video``
-- many documentation updates! (+ Colab demo Notebook update, general notebook updates)
-- set matplotlib to 3.0.3 and numpy to 1.14.5 in setup.py as latest versions cause errors.
-- splitting log files during training of multiple networks
-- cropping parameters can be directly passed to deeplabcut/pose_estimation_tensorflow/predict_videos.py
-- fixed bug for GPU allocation in deeplabcut/pose_estimation_tensorflow/predict_videos.py
-- optional manual frame grabber scroll bar extension 316

**New features:**
-- 3D project creation that includes 2 camera support for 3D analysis.
- These functions handle camera calibration, undistortion, checking alignment, and 3D triangulation with a new .h5 or .csv output file. It also includes tools to plot in 3D space alongside each 2D view.
-- new 3D_Demo_DeepLabCut.ipynb
-- new 3D documentation
-- new video tutorial for 3D: https://www.youtube.com/watch?v=Eh6oIGE4dwI

This is the python package of DeepLabCut. This package includes graphical user interfaces to label your data, and take you from data set creation to automatic behavioral analysis. It also introduces an active learning framework to efficiently use DeepLabCut on large experimental projects.

User guide on BioRxiv: https://www.biorxiv.org/content/early/2018/11/24/476531

This is the python package of DeepLabCut. This package includes graphical user interfaces to label your data, and take you from data set creation to automatic behavioral analysis. It also introduces an active learning framework to efficiently use DeepLabCut on large experimental projects.

User guide on BioRxiv: https://www.biorxiv.org/content/early/2018/11/24/476531

This is the python package of DeepLabCut. This package includes graphical user interfaces to label your data, and take you from data set creation to automatic behavioral analysis. It also introduces an active learning framework to efficiently use DeepLabCut on large experimental projects.

User guide on BioRxiv: https://www.biorxiv.org/content/early/2018/11/24/476531

- updated refinement GUI to novel version (thanks Tanmay!)

**Various bug fixes:**
- frame extraction error for uniform with start >0 https://github.com/AlexEMG/DeepLabCut/issues/190
- config files are organized in a structured way and now fully backwards compatible (e.g. https://github.com/AlexEMG/DeepLabCut/issues/191)
- fix and comment for too long file names (in Windows) https://github.com/AlexEMG/DeepLabCut/pull/195 and https://github.com/AlexEMG/DeepLabCut/issues/196
- fixed potential sorting problem with labeling GUI for some datasets https://github.com/AlexEMG/DeepLabCut/issues/193