Recently, the AISpeech-Shanghai Jiao Tong University (SJTU) X-LANCE Lab has achieved a significant academic milestone, with a total of 13 research papers accepted by two top-tier international conferences: ACL 2025 and INTERSPEECH 2025. ACL is an A-class conference per China Computer Federation (CCF) recommendations, representing the highest level in computational linguistics and natural language processing. INTERSPEECH, organized by ISCA, is the world's most comprehensive conference in speech signal processing.

The accepted papers cover cutting-edge research directions, including Retrieval-Augmented Generation (RAG) for LLM, speech codec and representation learning, low-latency speech synthesis, low-resource speech recognition, and reliable audio language models. These works make critical breakthroughs in multi-modal RAG, speaker-decoupled low-bitrate coding, efficient non-autoregressive speech synthesis, large-scale low-resource corpus construction with self-training, and model reliability enhancement. Collectively, they advance the development of efficient, reliable, and low-resource-friendly intelligent speech and language technologies, strengthening AISpeech's full-chain core technical capabilities and business innovation.

Highlighted research includes:

· NeuSym-RAG: Hybrid Neural Symbolic Retrieval with Multiview Structuring for PDF Question Answering: 

· LSCodec: Low-Bitrate and Speaker-Decoupled Discrete Speech Codec: 

· Unlocking Temporal Flexibility: Neural Speech Codec with Variable Frame Rate: 

This achievement underscores the lab's profound expertise and exploration in key AI speech and language technologies. AISpeech maintains a strong record of academic excellence, with numerous publications at premier conferences like ICASSP, EMNLP, AAAI, and ICML. The company is committed to tightly integrating research with industrial application, driving the practical transformation of scientific achievements.

As a dedicated conversational AI platform company, AISpeech possesses robust capability for foundational innovation and application development. It is recognized as a key player in the national AI strategy, operating the "MOST's National Platform for Language Computing AI Innovation", and has undertaken over ten national and ministerial-level projects.

AISpeech adheres to an independent LLM strategy, building a "1+N Reliable Distributed LLM Agent System" to serve enterprise clients effectively.