zhouwg / kantv

workbench for learing&practising AI tech in real scenario on Android device, powered by GGML(Georgi Gerganov Machine Learning) and NCNN(Tencent NCNN) and FFmpeg
Apache License 2.0
117 stars 19 forks source link

[whisper.cpp] sometimes some repeated sentences were shown in real-time English subtitle for English online-tv #84

Open zhouwg opened 5 months ago

zhouwg commented 5 months ago

whisper.cpp is an open-source and powerful/excellent/amazing device-side AI framework/lib/model for ASR(Automatic Speech Recognition, a sub-filed of AI).

I found there is a strange issue which some repeated sentences were shown in real-time English subtitle for English online-tv after finished integration work on Xiaomi14(although I think this PoC should also works fine on original powerful Google Pixel phone but I have no Google Pixel phone and can't validate that).

I'll try uploading a short video(it's a random issue) to illustrate this issue.

https://github.com/cdeos/kantv/assets/6889919/6d59e8e2-c33b-4db7-96f9-0c6176870305