Skip to content

Latest commit

 

History

History
129 lines (115 loc) · 10.2 KB

CHANGELOG.md

File metadata and controls

129 lines (115 loc) · 10.2 KB
Raw

OpenDR Toolkit Change Log

Version 3.0.0

Released on Dec, 4, 2023.

  • New Features:

    • Binary High Resolution Learner (#402)
    • ROS2 node for EfficientLPS (#404)
    • Fall and wave detection ROS nodes (#423)
    • Continual SLAM: Adds a new SLAM tool for Continual SLAM (#424)
    • Add RGB gesture recognition (#436)
    • FSeq2-NMS (#442)
    • Intent recognition tool (#443)
    • Robotti human detection simulation demo (#451)
    • Object Detection 2D Class Filtering (#467)
    • RL-based Learner for Active Face Recognition (#473)
    • YOLOv5s Inference Demo with Optimized Weights for Agricultural Use (#476)
    • Adaptive HR Pose Estimation (#479)

  • Enhancements:

    • Wave detection demo based on pose estimation (#394)
    • Facial expression recognition demo update (#405)
    • Object detection 2d camera demos (#408)
    • High Resolution Pose Estimation webcam demo (#409)
    • ROS nodes FPS performance measurements (#419)
    • Refactoring: pythonic joins in test_clang_format.py/test_cppcheck.py (#455)
    • Test-tools improvement (#456)
    • Adding prompt when transcribe with Whisper (#462)

  • Bug Fixes:

    • Fix package creator and sources (#390)
    • Lightweight OpenPose tool fixes and improvements (#392)
    • Fall Detection - alternative infer input (#397)
    • Yolov5 training bugfix (#401)
    • Fix the dependency conflict of geffnet installation (#410)
    • Fix bug in GEM ROS2 node (#420)
    • Fix link to nanodet documentation (#421)
    • EfficientLPS panoptic segmentation coloring bug (#426)
    • Bump flask from 1.1.2 to 2.3.2 (#430)
    • Fix tests on master branch (#438)
    • Added unzip installation as base ubuntu dependency and tool tests fixes (#454)
    • Active face recognition demo and bug fixes on Face Recognition (#459)
    • GPU installation fix (#463)
    • Fix ROS1 nodes argparse issue with .launch files (#465)
    • Minor fix on yolov5 webcam demo (#466)
    • Apply cuDNN init fix to all Object Detectors 2D (#469)
    • Updated test_suite_develop.yml based on latest test_suite.yml (#471)
    • Fix fmpgmapping (#472)
    • Synchronization and bugfixes (#478)

Version 2.2.0

Released on July, 3rd, 2023.

  • Dependency Updates (#431):
    • Updated PyTorch version from 1.9.0 to 1.13.1.
    • Updated Torchvision version from 0.10.0 to 0.14.1.
    • Updated Torchaudio version from 0.9.0 to 0.13.1.
    • Downgraded wheel version to 0.38.4 due to bug in recent version.

Version 2.1.0

Released on February, 22nd, 2023.

  • New Features:
    • Added Efficient LiDAR Panoptic Segmentation (#359).
    • Added Nanodet 2D Object Detection tool (#352).
    • Added C API implementations of NanoDet 2D Object Detection tool (#352).
    • Added C API implementations of forward pass of DETR 2D Object Detection tool (#383).
    • Added C API implementations of forward pass of DeepSORT 2D Object Tracking tool (#383).
    • Added C API implementations of forward pass of Lightweight OpenPose, Pose Estimator tool (#383).
    • Added C API implementations of forward pass of X3D 2D Activity Recognition tool (#383).
    • Added C API implementations of forward pass of Progressive Spatiotemporal GCN Skeleton-based Action Recognition tool (#383).
    • Added Binary High Resolution Analysis tool (#402).
    • Added Multi-Object-Search tool (#363)
  • Enhancements:
    • Added support in C API for detection target structure and vector of detections (#352)
    • Added support in C API for tensor structure and vector of tensors (#383)
    • Added support in C API for json parser (#383)

Version 2.0.0

Released on December, 31st, 2022.

  • New Features:
    • Added YOLOv5 as an inference-only tool (#360).
    • Added Continual Transformer Encoders (#317).
    • Added Continual Spatio-Temporal Graph Convolutional Networks tool (#370).
    • Added AmbiguityMeasure utility tool (#361).
    • Added SiamRPN 2D tracking tool (#367).
    • Added Facial Emotion Estimation tool (#264).
    • Added High resolution pose estimation tool (#356).
    • Added ROS2 nodes for all included tools (#256).
    • Added missing ROS nodes and homogenized the interface across the tools (#305).
  • Bug Fixes:
    • Fixed BoundingBoxList, TrackingAnnotationList, BoundingBoxList3D and TrackingAnnotationList3D confidence warnings (#365).
    • Fixed undefined image_id and segmentation for COCO BoundingBoxList (#365).
    • Fixed Continual X3D ONNX support (#372).
    • Fixed several issues with ROS nodes and improved performance (#305).

Version 1.1.1

Released on June, 30th, 2022.

  • Bug Fixes:
    • Fix Efficient Panoptic Segmentation submodule commit (#268).
    • Fix Face Recognition compilation error (#267).

Version 1.1.0

Released on June, 14th, 2022.

  • New Features:
    • Added end-to-end planning tool (#223).
    • Added seq2seq-nms module, along with other custom NMS implementations for 2D object detection.(#232).
  • Enhancements:
    • Added support for modular pip packages allowing tools to be installed separately (#201).
    • Simplified the installation process for pip by including the appropriate post-installation scripts (#201).
    • Improved the structure of the toolkit by moving io from utils to engine.helper (#201).
    • Added support for post-install scripts and opendr dependencies in .ini files (#201).
    • Updated toolkit to support CUDA 11.2 and improved GPU support (#215).
    • Added a standalone pose-based fall detection tool (#237)
  • Bug Fixes:
    • Updated wheel building pipeline to include missing files and removed unnecessary dependencies (#200).
    • panoptic_segmentation/efficient_ps: updated dataset preparation scripts to create correct validation ground truth (#221).
    • panoptic_segmentation/efficient_ps: added specific configuration files for the provided pretrained models (#221).
    • c_api/face_recognition: pass key by const reference in json_get_key_string() (#221).
    • pose_estimation/lightweight_open_pose: fixed height check on transformations.py according to original tool repo (#242).
    • pose_estimation/lightweight_open_pose: fixed two bugs where ONNX optimization failed on specific learner parameterization (#242).
  • Dependency Updates:
    • heart anomaly detection: upgraded scikit-learn runtime dependency from 0.21.3 to 0.22 (#198).
    • Relaxed all dependencies to allow future versions of non-critical tools to be used (#201).

Version 1.0

Released on December 31st, 2021.