12月15日，小米公司创始人、董事长、首席执行官雷军发文宣布，小米联合萨里大学、清华大学、海天瑞声联合发起第二届音频编码器能力挑战赛，将于明年9月同步亮相国际语音顶级会议 Interspeech 2026，目前已正式开放报名。 国际语音顶级会议 Interspeech 2026 将于明年 9 月在澳大利亚悉尼举行。由小米、萨里大学、清华大学、海天瑞声联合发起的第二届 Audio Encoder ...

这项由马里兰大学和Meta公司联合完成的突破性研究发表于2025年5月28日的arXiv预印本平台（arXiv:2505.22664v1 [cs.CV]），论文题为《通过LLM替身实现零样本视觉编码器嫁接》(Zero-Shot Vision Encoder Grafting via LLM Surrogates)。该研究由Kaiyu Yue、Vasu Singla、Menglin ...

SK Telecom has unveiled a universal document interpretation technology for vision-language model (VLM) and large language model (LLM) training, based on its proprietary large language model, A.Dot X ...

Transformer-based models have rapidly spread from text to speech, vision, and other modalities. This has created challenges for the development of Neural Processing Units (NPUs). NPUs must now ...

Apple has announced its own visual language model (VLM), ' FastVLM '. Conventional VLMs have the problem of decreasing efficiency as their accuracy increases, but FastVLM maintains high accuracy while ...