Openai-game-translator

Release for final evaluation
- Added a small example in README.
- Support fast, accurate AWS (live) and AWS xunfei (prerecorded) transcription.
- Support fast and accurate OpenAI translation using gpt3.5-turbo.

Todo
- Enhance AWS_pre and xunfei performance without prerecorded audio files.
- Add and save context when translating a game.
- Implementing conversational audio translation.

- Improved precision with `aws_live` transcription.
- Discard `sounddevice` by `pyAudio` to capture the audio stream.
- Adjusted recording params (`MAX_LOW_AUDIO_FLAG`, `AUDIO_MIN_RMS` etc.).
- Visualized recording volume.
- Modified shell input arg parser, enabled subcommand and parameter selection with mutual constraints.
- Enabled multiple target and input language selection, now supporting English, Chinese, French, German, Japanese, Korean, Russian and Spanish.
- Purified code.

Current support English <-> Chinese translation with aws prerecorded/aws live/xunfei speed transcription. Currently, support audio detection based on self-defined ams calculation, will upgrade to more fluent detection.