laitimes

"National AI Carnival" will be staged on October 24, and the reporter will visit the museum in advance to give you a glimpse of the wonderful exhibits

author:Dawan News

Dawan News Xunfei Xinghuo Cognitive Big Model V3.0 was officially released, various robots such as humanoid robots took turns to "grab the camera", and the first "Science and Art Exhibition" showed the perfect combination of technology and art... From October 24th to 29th, the 6th World Sound Expo and 2023 Global 1024 Developer Festival will be held in Hefei Sports Center, which will bring you a "National AI Carnival". On October 23, reporters from Xin'an Evening News and Dawan News visited the museum in advance to see the wonderful exhibits for readers and netizens.

"National AI Carnival" will be staged on October 24, and the reporter will visit the museum in advance to give you a glimpse of the wonderful exhibits

Humanoid robots and other "appearances"

At last year's World Sound Expo and Global 1024 Developer Festival, the quadruped robot brought by iFLYTEK could jump and jump up, up and down, which became one of the focuses of the exhibition area. At this year's conference, as the latest achievement of the "iFLYTEK Super Brain 2030" plan, the industry's first humanoid robot with integrated large model and embodied intelligence and its home companion robot debuted, attracting more attention.

The reporter saw at the scene that the robot has the shape of a human, two hands and two feet. In complex terrain areas with slopes and grass, it can walk autonomously on its legs. According to reports, this is the industry's first system-level demo of "integrating domestic large language model + preliminary embodied intelligence + humanoid robot", "In the field of robotics, it is a pull through the real sense of realizing general artificial intelligence and general robots." "Staff introduction.

Humanoid robots can not only walk autonomously, but also perform simple movements. It turns out that it has a "brain" that can execute open instructions and tasks through large model semantic understanding and behavior planning, and complete more flexible embodied intelligent interactions. At the same time, based on reinforcement learning and AI motion control, it can move autonomously to find objects, and interact with them to grasp, place and other operations.

In the Science and Technology Museum in the exhibition area of Hefei Sports Center, there are many different robots, including companion robots, chess robots, calligraphy robots, etc., and visitors can fully experience the magic of "AI robots".

"National AI Carnival" will be staged on October 24, and the reporter will visit the museum in advance to give you a glimpse of the wonderful exhibits

The Spark cognitive large model occupies the C position

In the 6th World Sound Expo and 2023 Global 1024 Developer Festival, iFLYTEK Xinghuo cognitive large model occupies the C position. In the "1+N" experience area of the iFLYTEK Spark cognitive model of the Science and Technology Museum, various technical capabilities upgraded by iFLYTEK Spark V3.0 are displayed in a concentrated manner, such as multi-modal, multilingual voice, code, reasoning capabilities, etc., as well as a number of practical and interesting assistants in iFLYTEK Spark.

"National AI Carnival" will be staged on October 24, and the reporter will visit the museum in advance to give you a glimpse of the wonderful exhibits

Here is the generation of articles, you can send a request, iFLYTEK Spark can generate articles and stories that meet your requirements; There is also a picture generation, you want the machine to draw a picture of you climbing the snow mountain on the spot, iFLYTEK Spark can complete it immediately; In addition, here, you can also generate a "virtual self" in the 3D world, you can take a photo, record a voice, and the machine can quickly complete the personalized generation and drive of the 3D virtual person. "Personalized 3D virtual humans involve a number of key technologies, including 3D face reconstruction technology based on 3D data-driven 3D data generation, personalized speech synthesis technology based on a small number of speaker audio, and voice-driven face expression generation technology."

In addition, there are AI debate competitions, multilingual assistants, etc., allowing you to experience the endless charm of artificial intelligence cognitive large models.

"National AI Carnival" will be staged on October 24, and the reporter will visit the museum in advance to give you a glimpse of the wonderful exhibits

"Xiaoyi" health assistant was officially released

After five years of development, the World Sound Expo and 2023 Global 1024 Developer Festival are gradually becoming an artificial intelligence event with international influence. The 2023 conference is held in the wave of explosive development of artificial intelligence cognitive large models. How can cognitive big models empower thousands of industries and make them smarter? In the conference exhibition area, you can also find the answers.

Among them, the medical big model and iFLYTEK Xiaoyi App officially released at this conference are based on medical big model technology, covering medical scenarios such as disease prevention, disease self-examination, report interpretation, medical information query, and TCM differentiation, and are committed to assisting users to improve medical efficiency and medical experience, and become everyone's AI health assistant. "It has functions such as self-screening, report analysis and guidance, medical information query and recommendation, etc., which can provide users with convenient health services." The staff said.

In addition, the iFLYTEK Spark Legal Model was also launched at the conference to create a "smart brain" in the field of legal technology. According to the staff, the legal model focuses on four technical directions: knowledge question and answer, content generation, language understanding, and logical reasoning, and customizes the development of 12 industry capabilities. For example, it can provide case-handling auxiliary functions such as case analysis and document assistance generation. "In the process of reading the case files and documents in the traditional mode, judges and prosecutors need to analyze and determine the legal facts one by one on the materials of the entire file, but relying on the Spark model, the content of the file can be automatically extracted from all elements, quickly sort out the facts of the case, and summarize the focus of the dispute in the case to the judge handling the case for confirmation, and in the document generation stage, automatically generate the content of the judgment document through the results of the fact determination and based on the legal basis." Staff introduction.

"National AI Carnival" will be staged on October 24, and the reporter will visit the museum in advance to give you a glimpse of the wonderful exhibits

Acoustic imager just "returned" from the Asian Games Village

In the industrial pavilion of the exhibition area, the Antelope industrial large model with five core capabilities, the industrial six-sense robot with "ears and eyes" and "agility" are displayed, and the "black light" laboratory of intelligent inspection in the field of intelligent interaction are displayed... Here, you can see various industrial hardcore black technologies and experience the key achievements of iFLYTEK in the deep integration of new generation information technology and industry.

At the recent Hangzhou Asian Games, a iFLYTEK acoustic imager was "put to work" in the Asian Games Village, which successfully ensured the safety of gas use in the Asian Games. At this sound expo, the heavily upgraded iFLYTEK acoustic imager also came. "Common industrial faults, under the influence of many interference factors, it is difficult to locate the specific location of the fault by manpower alone. With this DP300, the location of faulty sound sources is 'clear at a glance'. The upgraded 132 microphone acoustic imager has higher detection accuracy. "Experiments have proved that the acoustic imager can find multiple leaks such as flanges and valves within 30 minutes and achieve 100% accuracy, compared with 5 hours if the traditional artificial soapy water method is used.

Last year, iFLYTEK's intelligent four-legged robot dog "Xiaohei" earned enough eyeballs. This time, it came to the scene with the industrial six-sense robot "legion". Robot dogs, drones and robotic arms that integrate the six senses of industry can not only cope with complex terrain operations, but also realize high-altitude inspection such as high-voltage poles and ultra-high-voltage lines, and build an industrial six-sense intelligent inspection with "space-ground integration". "Simulate experienced workers, with the ability to 'listen, smell, smell and touch', and can focus 24 hours to solve industrial production problems such as equipment management, safety testing, energy conservation and carbon reduction... At present, it has been applied in electric power, mining, steel, chemical industry, wind power, hydropower and other fields. ”

"National AI Carnival" will be staged on October 24, and the reporter will visit the museum in advance to give you a glimpse of the wonderful exhibits

"Symphony of the Sun" was presented at the Science and Art Exhibition

The 6th World Sound Expo and 2023 Global 1024 Developer Festival were held in Hefei Sports Center located in the government affairs district, with science and technology halls, industrial halls, industry halls, education halls, life halls, ecological halls and science and art exhibitions, displaying hundreds of applications under the deep empowerment of large models.

In the industry pavilion, it is divided into three modules: smart city, financial technology and operator; The Education Hall has six exhibition areas: smart teaching, science education, language learning, physical and mental health, happy growth and independent learning; The Life Pavilion exhibits intelligent products and applications in five scenarios: home, health, learning, office and travel; The ecological pavilion includes five sections: iFLYTEK Open Platform, Lingsi Technology AI Chip, Industry Acceleration Center, HKUST Silicon Valley, and Anhui Artificial Intelligence Association, fully displaying the open platform ecology that gathers 5.4 million developers.

In addition, the first China (Hefei) International Science and Art Festival is a scientific and technological cultural event, aiming to showcase the perfect combination of science and art, and promote scientific and technological innovation and cultural exchanges. The science and art exhibition here brings together the works of top scientists, artists, experts and scholars at home and abroad, and allows everyone to experience the beauty of the integration of science and technology and art through unique science and art exhibition areas such as "Map of the History of Technology", Big Science Installation Art Documentary Exhibition, and Science and Technology Culture Interactive Experience Exhibition.

Among them, "Map of the History of Technology" is an ink on paper art installation created by artist Qiu Zhijie for more than ten years on the basis of integrating traditional Chinese ink calligraphy and emerging digital technology. Solar Symphony is the generative art of the fully superconducting tokamak Oriental Hyperring Installation (EAST). The 3.3D Rubik's Cube exhibit is spliced on three square screens to form a cube side to construct a square three-dimensional structure that can carry the picture, presenting the naked-eye 3D visual effect with a fixed viewing angle.

"National AI Carnival" will be staged on October 24, and the reporter will visit the museum in advance to give you a glimpse of the wonderful exhibits

Dawan News reporter Xiang Lei Liu Yang

Edited by Wang Cui