..
  # SPDX-FileCopyrightText: Copyright (c) 2022-2023 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
  # SPDX-License-Identifier: Apache-2.0
  #
  # Licensed under the Apache License, Version 2.0 (the "License");
  # you may not use this file except in compliance with the License.
  # You may obtain a copy of the License at
  #
  # http://www.apache.org/licenses/LICENSE-2.0
  #
  # Unless required by applicable law or agreed to in writing, software
  # distributed under the License is distributed on an "AS IS" BASIS,
  # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
  # See the License for the specific language governing permissions and
  # limitations under the License.

.. _v0.3.1-beta:

Beta.1
======

The v0.3.1 release provides several bug fixes along with documentation updates and performance improvements.

Release Highlights
------------------

* Added documentation for Python APIs
* Added CentOS 7 build support with GLIBC >= 2.17
* Improved APIs with C handler reference counting, C++ move semantics, and VarShape iterators
* Removed legacy interfaces typedefs: IImage, ITensor, and IImageBatch
* Handler Manager no longer requires resource requirements object
* Moderate performance improvement for Non-Maximal Suppression (NMS) operator

Bug Fixes
---------

* Addressed `QUESTION #23 <https://github.com/CVCUDA/CV-CUDA/issues/23>`_: RuntimeError: cudaErrorIllegalAddress: an illegal memory access was encountered`
* Addressed `BUG #72 <https://github.com/CVCUDA/CV-CUDA/issues/72>`_: RuntimeError: pybind11_object_dealloc(): Tried to deallocate unregistered instance!
* Addressed `BUG #77 <https://github.com/CVCUDA/CV-CUDA/issues/77>`_: Memory sanitizer detects multiple problems when running nvcv_test_types_python
* Fixed Pillow Resize operator bug in clip8_lookups


Known Issues
------------

- Morphology Operator: poor benchmarking performance when iteration parameter > 1
  - Workaround: Set iteration parameter == 1 and run operator multiple times
  - Fix expected in the next release

Acknowledgements
----------------
CV-CUDA is developed jointly by NVIDIA and the ByteDance Machine Learning team.