refactor: update hallucination detection logic and improve repetition threshold

feat: implement context-aware pronoun resolution in translation process

feat: enhance translation logic with context-aware handling and improved fallback mechanisms

fix: remove unnecessary comment about transformers dependency in requirements.txt

feat: update translation dependencies to use HuggingFace Transformers for improved performance

refactor: streamline lifespan management and enhance user session history handling

feat: replace ArgosTranslator with HelsinkiTranslator for improved translation accuracy and context support

feat: implement HelsinkiTranslator for improved translation capabilities with model loading and pivot support

fix: change default model size from medium to base in WhisperASR initialization

fix: increase idle wait time in _read_utterance method for improved processing

fix: increase pipeline timeout from 30s to 300s for user sessions

fix: reduce pipeline timeout from 300s to 30s and handle self-captioning for connected users

fix: increase timeout for reading utterances and adjust idle wait timing

feat: add root endpoint for backend status and documentation links

fix: add display name handling for user sessions in voice bridge

fix: restore missing language pairs for translation in ArgosTranslator

fix: update ASR model size to medium for improved performance

fix: update model directory initialization to use environment variable

fix: make server port configurable via environment variable

fix: add Chinese voice model to model map in PiperTTS

fix: remove wave from requirements (stdlib module)

docs: comprehensive README with HF Spaces config

feat: Add endpoint to check room existence and occupant count

feat: Update WhisperASR model size to medium and configure ThreadPool settings for improved performance

feat: Increase ThreadPoolExecutor workers and adjust VAD parameters for improved performance

feat: Enhance TTS shutdown process and improve VAD parameters for better audio handling

feat: Add centralized logging setup with file and console handlers

feat: Increase max speech duration in StreamingVAD to enhance processing capabilities

feat: Update ASR model size to medium and clean up warmup pairs in TranslationPipeline

feat: Refactor TTS model for improved model selection, error handling, and process management

feat: Enhance hallucination detection and repetition loop handling in ASR model

feat: Update requirements to include faster-whisper for ASR enhancements

feat: Enhance voice call translation server with improved user session management, WebSocket handling, and pipeline timeout adjustments

feat: Update StreamingVAD with enhanced calibration guide, improved energy threshold handling, and refined logging for speech detection

feat: Improve TranslationPipeline with enhanced logging, error handling, and code clarity

feat: Refactor ASR model to use faster-whisper for improved performance and remove subprocess dependencies

feat: Enhance TTS model with startup verification for missing models, improve subprocess handling, and add error messages for missing Piper installation

feat: Enhance voice processing with improved VAD, ASR, and TTS models; update WebSocket connection for production

feat: Add webrtcvad to requirements for improved voice activity detection

feat: Add CORS middleware to allow cross-origin requests for the FastAPI application

feat: Enhance voice processing with Streaming VAD and improve audio handling

feat: Implement Voice Activity Detection (VAD) for automatic speech chunking

feat: Initial commit of Voice Translation App with backend and frontend setup

feat: Initial commit of Voice Translation App with backend and frontend setup