Search Issue Tracker
Won't Fix
Won't Fix in 2023.2.X
Votes
0
Found in
2020.3.42f1
2021.3.18f1
2022.2.5f1
2023.1.0b2
2023.2.0a14
Issue ID
UUM-25512
Regression
No
Performance issues when instantiating a recognizer under UnityEngine.Windows.Speech
Reproduction steps:
- Open the attached project “RecognizerIssue”
- In the project, navigate to "Assets/Scenes" and open "Sample Scene"
- Open the profiler (Window → Analysis → Profiler)
- In the CPU Usage module disable everything except Scripts
- Enter Play Mode
Expected result: Performance stays consistent
Actual result: Around frame 100 there is a large spike in the CPU usage section of the profiler
Reproducible with: 2020.3.42f1, 2021.3.18f1, 2022.2.5f1, 2023.1.0b2
Reproducible on: Windows 10
Add comment
All about bugs
View bugs we have successfully reproduced, and vote for the bugs you want to see fixed most urgently.
Latest issues
- Texture generator reference image fails despite valid size
- Agent mode can enter infinite execution loop
- AI Gateway responses contain duplicated sentences
- Texture 2D asset is created when promoting a generated Cubemap to a new asset
- Unable to copy multiple paragraphs from Assistant responses
Resolution Note:
Thank you for your bug report.
We are not able to improve the performance of the API provided by Unity without a major internal rewrite. After reviewing the issue and the impact it has on our users our team has decided that this case will be resolved as "Won't Fix". This does not mean that the issue isn’t a legitimate bug, but instead that we are not able to prioritize the fix, at this time.
A workaround is available on the UWP platform. You can use, from a background thread, the Windows.Media.SpeechRecognition.SpeechRecognizer class from the Windows Runtime APIs.
For more information on using Windows Runtime APIs in Unity, see https://docs.unity3d.com/Manual/IL2CPP-WindowsRuntimeSupport.html. For more information on using the SpeechRecognizer API in C#, see https://learn.microsoft.com/en-us/windows/apps/design/input/speech-recognition.
The case will now be closed, and will not be reopened unless new information arises that would change the issue’s impact. Please let us know if you have additional information relating to the severity of this bug.
Resolution Note (2023.2.X):
Thank you for your bug report.
We are not able to improve the performance of the API provided by Unity without a major internal rewrite. After reviewing the issue and the impact it has on our users our team has decided that this case will be resolved as "Won't Fix". This does not mean that the issue isn’t a legitimate bug, but instead that we are not able to prioritize the fix, at this time.
A workaround is available on the UWP platform. You can use, from a background thread, the Windows.Media.SpeechRecognition.SpeechRecognizer class from the Windows Runtime APIs.
For more information on using Windows Runtime APIs in Unity, see https://docs.unity3d.com/Manual/IL2CPP-WindowsRuntimeSupport.html. For more information on using the SpeechRecognizer API in C#, see https://learn.microsoft.com/en-us/windows/apps/design/input/speech-recognition.
The case will now be closed, and will not be reopened unless new information arises that would change the issue’s impact. Please let us know if you have additional information relating to the severity of this bug.