Deeplabcut

Latest version: v2.3.11

Safety actively analyzes 710445 Python packages for vulnerabilities to keep your Python projects secure.

Scan your dependencies

Page 7 of 11

2.1.6

- auto-check video quality before use (i.e. is it corrupt?) 558. 561
- utility function: extra catch if config file not found 578
- allows crop coordinates to be read from config.yaml in main GUI when analyzing videos
- minor GUI minor updates (small bug fixes) 584, 564
- set numpy to 1.16.4 in setup.py to avoid warnings when loading with 1.17.2
- allow tables to be most recent (compatibility with python 3.7* but main GUI users should continue to use python 3.6)
- doc updates 589

2.1.5.2

Actually includes these two parts (was missing in DLC 2.1.5)
- Accelerate training set creation (by at least 2x, up to 10x under multiple train size/split conditions);
- Possibility to pass lists of train–test indices for user-defined splits in create_training_dataset()
- Fix to https://github.com/AlexEMG/DeepLabCut/pull/555 (related to #550 552 553)
- Includes https://github.com/AlexEMG/DeepLabCut/pull/551 (conversion code update for new scorer name convention)
- Expanded installation guide for MacOs

2.1.5

Happy new year! This version has a few minor updates. A big release is around the corner...

- Accelerate training set creation (by at least 2x, up to 10x under multiple train size/split conditions);
- Possibility to pass lists of train–test indices for user-defined splits in create_training_dataset()
- Selecting bodyparts for plotting; fixed https://github.com/AlexEMG/DeepLabCut/issues/504
- plot_trajectories now allows for selection of bodyparts (also in the Project GUI)
- create_labeled_videos now allows selection of bodyparts in Project GUI
- fix for 3D plotting if videos of different views have different length https://github.com/AlexEMG/DeepLabCut/commit/de35651835f8e3e59d09b2533fa6f88ab64ff449
- reading num_outputs from configfile (if it exists): https://github.com/AlexEMG/DeepLabCut/commit/2fe0fccd904b95ead21cfb2bfa0591e4ef26f996
- Better message when no videos are found during analysis: https://github.com/AlexEMG/DeepLabCut/commit/d275c3cf91d16a5d122814894b8aec711ec47af9

This release includes contributions from the new DLC-ChanZuckerbergInitiative fellows JessyLauer jeylau, alexEMG and MMathislab

2.1.4

- new command to start the GUI: python -m deeplabcut
https://github.com/AlexEMG/DeepLabCut/pull/484
- updated docs with regard to types of augmentation available and other aspects
- cleaned up loaders (now will throw out corrupt images; training more stable) https://github.com/AlexEMG/DeepLabCut/issues/463
- fix for negative coordinates https://github.com/AlexEMG/DeepLabCut/pull/501

2.1.1

Bug fixes and minor edits:
- fix docstring in tensorpack and calculation of processratio https://github.com/AlexEMG/DeepLabCut/pull/455
- multi peak extraction had an error (in 2.1 but not before): https://forum.image.sc/t/issue-with-analyzing-multiple-body-parts/30453/6
- one can pick more parameters in the GUI

2.1

**New Features, major updates and upgrades:**
- new **Project Manager GUI** (see docs/PROJECT_GUI.md)
- **dynamic auto-cropping** is available (issue 198): use during analyze_videos: deeplabcut.analyze_videos(path_config_file,[video], dynamic=(True,.1,30)). This dramatically improves the inference speed for typical applications in neuroscience such as open field behaviors where the animal is much smaller than the frame and no prior tracking is necessary.
- Integrated **MobileNetV2s** (described in Pretraining boosts out-of-domain robustness for pose estimation by Alexander Mathis, Mert Yüksekgönül, Byron Rogers, Matthias Bethge, Mackenzie W. Mathis https://arxiv.org/abs/1909.11229)
- integrated **TF based inference** code (based on https://arxiv.org/abs/1909.11229).
- Added in new data augmentation options with imgaug see Box 2 in docs (https://github.com/aleju/imgaug)
- New function, create_training_model_comparison that allows the user to create identical shuffles for multiple network and/or data augmentation testing.
- Updated evaluation metric: one can now evaluate the model at different scales (by passing rescale=True one can evaluate the model at the 'global_scale' variable (as set in the test/pose_config.yaml file for a particular project). I.e. every image will be resized according to that scale and prediction will be compared to the resized ground truth. The error will be reported in pixels at rescaled to the *original* size. I.e. For a [200,200] pixel image evaluated at global_scale=.5, the predictions are calculated on [100,100] pixel images, compared to 1/2*ground truth and this error is then multiplied by 2. The evaluation images are also shown for the original size!
- Retraining now also loads deconvolution weights by default
- dlc scorer name changed, and backwards compatibility maintained. E.g. data for videos like abc.mp4 will be saved as abcDLC_resnet50_xyz.h5 or abcDLC_mobnet35_xyz.h5 (depending on the network)

**Bug fixes and minor edits:**
- Changed imresize and imread from scipy.misc (deprecated functions); now uses openCV
- Colors changed when manually extracting outliers https://forum.image.sc/t/extract-outlier-frames-saves-images-with-different-color/29868/2
- Minor edits to ShortenVideo function, now returns path to shortened video
- By default .csv exported when analyzing a video when initializing a human pre-trained network (https://github.com/AlexEMG/DeepLabCut/pull/427)
- Changes to default values in augmentation with TensorPack; See results here: https://github.com/AlexEMG/DeepLabCut/pull/426 > Thanks katierupp! and the from #429 thanks ppwwyyxx!
- Updated link to human DEMO on COLAB
- updated setup.py dependencies (addresses 434, 433, 441, 440)
- edited and expanded test functions
- expanded docs, updated wiki

Contributors: alexEMG, meet10may, MMathisLab, mertyg & tbiasi

Page 7 of 11

© 2025 Safety CLI Cybersecurity Inc. All Rights Reserved.