https://speech-expert.ru/news/geometric-audio-codec-resynthesis Speech Expert ru 2026-08-21T00:00:00Z Geometric Iterative Retrieval for Neural Audio Codec Resynthesis https://speech-expert.ru/news/sjepa-gmm-soft-target-mapping Speech Expert ru 2026-08-21T00:00:00Z Does Mapping Non-Maximal Probabilities to GMM Components Matter for S-JEPA Encoder Representations? https://speech-expert.ru/news/multilingual-medical-asr-layer-analysis Speech Expert ru 2026-08-21T00:00:00Z Understanding Multilingual Medical ASR Adaptation Through Layer-Wise Analysis https://speech-expert.ru/news/x2streaming-tts-token-level Speech Expert ru 2026-08-21T00:00:00Z X2Streaming-TTS: Causal Token-Level Text-to-Speech from Streaming Text with Speech-State Inheritance https://speech-expert.ru/news/loihi2-acoustic-anomaly-detection Speech Expert ru 2026-08-21T00:00:00Z Low-Power, Neuromorphic, Acoustic Anomaly Detection for Persistent Machine Monitoring https://speech-expert.ru/news/alignment-only-audio-language-model Speech Expert ru 2026-08-21T00:00:00Z Alignment Is All You Need: Instruction-Free Training for General Audio-Language Models