https://speech-expert.ru/news/hybrid-real-complex-phase-aware-speech-enhancement Speech Expert ru 2026-08-14T00:00:00Z Hybrid Real- and Complex-Valued Neural Network Architecture for Speech Enhancement https://speech-expert.ru/news/rt-semamba-realtime-speech-enhancement-distillation Speech Expert ru 2026-08-14T00:00:00Z RT-SEMamba: Real-Time Speech Enhancement Mamba via Progressive Knowledge Distillation https://speech-expert.ru/news/six-codec-latent-generative-speech-enhancement-paradigms Speech Expert ru 2026-08-14T00:00:00Z Rethinking Language Model-Based Generative Speech Enhancement in the Latent Space of a Neural Audio Codec https://speech-expert.ru/news/slt-2026-smartglasses-egocentric-speech-benchmark Speech Expert ru 2026-08-14T00:00:00Z The SLT 2026 SmartGlasses Challenge: Benchmarking Egocentric Multi-Talker Speech Recognition and Understanding with Audio-Language Models https://speech-expert.ru/news/on-policy-self-distillation-multi-dialect-asr Speech Expert ru 2026-08-14T00:00:00Z On-Policy Self-Distillation for Multi-Dialect ASR: Mastering Dialects, Retaining Mandarin https://speech-expert.ru/news/midashenglm-gen-unified-audio-scenes Speech Expert ru 2026-08-14T00:00:00Z MiDashengLM-Gen: Unified Audio Scene Generation via LLM-Driven Autoregressive Flow Matching https://speech-expert.ru/news/uniswap-streaming-audio-visual-identity Speech Expert ru 2026-08-14T00:00:00Z UniSwap: Streaming Audio-Visual Identity Swapping for Talking Videos https://speech-expert.ru/news/phoenix-tts-joint-tokenizer-flow-matching Speech Expert ru 2026-08-14T00:00:00Z Phoenix TTS: High-Fidelity Synthesis and Voice Conversion via Flow-Matching-Driven Speech Tokenization https://speech-expert.ru/news/confucius4-tts-transcript-free-cross-lingual Speech Expert ru 2026-08-14T00:00:00Z Confucius4-TTS: Transcript-Free Cross-Lingual Zero-Shot TTS with a Learnable Speaker Encoder https://speech-expert.ru/news/easper-accessible-asr-language-documentation Speech Expert ru 2026-08-14T00:00:00Z Easper: An Accessible ASR Pipeline for Language Documentation https://speech-expert.ru/news/relative-transfer-matrix-multisource-microphones Speech Expert ru 2026-08-14T00:00:00Z Deep Learning Based Relative Transfer Matrix Estimation for Multiple Sources and Multiple Microphones https://speech-expert.ru/news/luna-tts-diffusion-family-realtime Speech Expert ru 2026-08-14T00:00:00Z Luna-TTS Family Technical Report https://speech-expert.ru/news/cookvoice-controllable-speech-singing-generation Speech Expert ru 2026-08-14T00:00:00Z CookVoice: Unified Framework for Style Controllable Multi-Modal Human Voice Generation https://speech-expert.ru/news/infant-audio-whisper-structured-speaker-conditioning Speech Expert ru 2026-08-14T00:00:00Z Robust Multi-Tier Infant-Centered Audio Understanding with Whisper via Structured Speaker Conditioning https://speech-expert.ru/news/donorrank-language-selection-low-resource-asr Speech Expert ru 2026-08-14T00:00:00Z DonorRank: Donor Language Selection for Low-Resource Cross-Lingual Speech Recognition https://speech-expert.ru/news/qwen-musicavqa-7b-temporal-audio-tokens Speech Expert ru 2026-08-14T00:00:00Z Qwen-MusicAVQA-7B: A Multimodal Model for Music Audio-Visual QA https://speech-expert.ru/news/relfx-relative-audio-effects-representation Speech Expert ru 2026-08-13T00:00:00Z RelFx: относительные представления аудиоэффектов без dry reference https://speech-expert.ru/news/bioacoustic-denoising-time-frequency-ridge-tracking Speech Expert ru 2026-08-13T00:00:00Z Bioacoustic denoising: частотно-временное восстановление ультразвуковых вокализаций https://speech-expert.ru/news/dino-a-audio-self-distillation Speech Expert ru 2026-08-13T00:00:00Z DINO-A: отрицательный результат для самодистилляции аудио https://speech-expert.ru/news/garhwali-asr-multiseed-evaluation Speech Expert ru 2026-08-13T00:00:00Z Garhwali ASR: почему пять seed важнее новой функции потерь https://speech-expert.ru/news/voxsumm-speech-summarization-translation Speech Expert ru 2026-08-13T00:00:00Z VoxSumm: длинная устная новость, перевод и краткое изложение https://speech-expert.ru/news/never-stop-speaking-dos-speech-language-models Speech Expert ru 2026-08-13T00:00:00Z Never Stop Speaking: DoS-риск для речевых языковых моделей https://speech-expert.ru/news/duplexworld-voice-agent-benchmark Speech Expert ru 2026-08-13T00:00:00Z DuplexWorld: голосовые агенты звучат хорошо, но задачи решают не всегда https://speech-expert.ru/news/asr-roundtrip-masked-tts-reading-errors Speech Expert ru 2026-08-13T00:00:00Z ASR-roundtrip: как распознавание скрывает ошибки китайского TTS https://speech-expert.ru/news/x2-turn-streaming-asr-turn-state Speech Expert ru 2026-08-13T00:00:00Z X2-Turn: потоковое распознавание речи вместе с состоянием реплики https://speech-expert.ru/news/whisper-aware-llm-whispered-speech-recognition Speech Expert ru 2026-08-13T00:00:00Z Whisper-Aware LLM: распознавание шепотной речи без галлюцинаций https://speech-expert.ru/news/age-aware-child-speech-phoneme-recognition Speech Expert ru 2026-08-13T00:00:00Z Age-aware phoneme recognition: детская речь на устройстве https://speech-expert.ru/news/mymediwhisper-burmese-medical-asr Speech Expert ru 2026-08-13T00:00:00Z myMediWhisper: бирманская медицинская речь и адаптация Whisper https://speech-expert.ru/news/voice-anonymization-worst-case-privacy-disclosure Speech Expert ru 2026-08-13T00:00:00Z Worst-case privacy disclosure: почему EER мало для анонимизации голоса https://speech-expert.ru/news/bitse-binaural-target-speaker-extraction Speech Expert ru 2026-08-13T00:00:00Z BiTSE: бинауральное выделение целевого диктора для AR-очков