iFLYTEKThe company announced that it will launch iFlytek on June 27, 2024 at the Beijing National Convention Center.Spark Large ModelThe conference was themed "AI Assistant that Understands You" and released iFlytek Spark Big Model V4.0 and related landing applications.
Official disclosure of the press conference content:
-
Comprehensively improve the seven core capabilities of the large model base, benchmarking GPT-4 Turbo;
-
Newly upgraded iFlytek Spark App/Desk, released "Personal Space", and created an AI assistant that understands you better;
-
The newly released Spark Smart Review Machine helps teachers teach students in accordance with their aptitude; the newly upgraded iFlytek AI Learning Machine provides personalized learning assistance with AI 1-on-1 Q&A tutoring;
-
The iFlytek Xiaoyi App has been completely upgraded to create “everyone’s health assistant”;
-
Spark's voice big model capabilities have been further improved, with the release of multi-language/multi-dialect voice recognition without switching, multi-modal recognition in complex scenarios, voice interaction technology, and applications in automobiles, offices, and other products;
-
Released the Spark Enterprise Intelligent Agent Platform, and also released intelligent agent demonstration applications such as Business Opportunity Assistant and Procurement Assistant;
-
It provides cloud-edge-end and hardware-software integrated large model solutions to support the implementation of large models in various fields; and launches the new robot super brain platform 2.0 to promote the development of the robotics industry with a hardware-software integrated technology platform.
Liu Qingfeng, Chairman of iFLYTEK, revealed at the company's 25th anniversary celebration:iFlytek Spark V4.0 will showcase the latest end-to-end intelligent voice implementation resultsAt the same time, the company’s end-to-end voice technology modeling solution is also planned and is expected to be launched on the market in the third quarter.
In addition, iFLYTEK will also provide sentence replication, high-noise scene speech recognition, and multi-dialect and multi-language free switching functions that GPT-4o does not have.