The Paris Olympics have been in full swing since July 26. As a premier international stage representing fairness and justice, every score in the Olympics holds significant importance for the athletes’ futures. With the advancement of AI technology, AI is playing an increasingly vital role in assisting with event judgments and decisions.
Alibaba Cloud: AI Creates “Bullet Time” Viewing Experience
Alibaba Cloud has introduced innovative AI-enhanced technology that not only improves the quality of match footage but also contributes to the fairness of judgments. The high-degree-of-freedom replay “Bullet Time” technology provided by Alibaba Cloud uses an array of high-speed cameras and advanced cloud computing technology to achieve real-time capture and multi-angle reconstruction of match moments. When athletes glide into a specific area, the camera quickly selects the frame of the exciting moment and transmits it to the cloud for 3D modeling and rendering. Through effects such as freezing, rotating, and zooming, the details of the athletes’ movements are clearly displayed, providing an immersive viewing experience.
“Bullet Time” From Alibaba
Potential AI Technologies Involved:
– Computer Vision:Capturing game moments through high-definition cameras.
– 3D Modeling and Rendering:Conducting 3D modeling in the cloud to generate dynamic effects like freeze-frames, rotations, and zoom-ins.
– Data Visualization:Providing high-quality visual presentations for both audiences and referees to assist in making more accurate judgments.
Omega: AI-Powered Strobe Analysis
Omega, in collaboration with the Olympic Broadcasting Services (OBS), utilizes AI technology to provide services that are faster, more accurate, and more insightful. For instance, intelligent strobe analysis for sports such as tennis, diving, and track and field helps audiences better understand the athletes’ movements and biomechanical conditions. This not only enriches the viewing experience but also provides referees with more precise assistance for making judgments.
“Intelligent Strobe Analysis”From Omega
Potential AI technologies involved
1. Intelligent Video Analysis: Conducting strobe analysis for competitions such as diving, track and field, and rhythmic gymnastics.
2. Action Recognition and Analysis: Using deep learning algorithms to identify and analyze athletes’ movements and biomechanical states.
3. Enhanced Data Graphics: Utilizing AI to generate detailed data of athletes in the air and during water entry.
4. Motion Tracking Technology: Real-time monitoring and analysis of athletes’ positions and states, providing precise motion tracking data.
Intel: AI Precisely Captures Exciting Moments
Intel’s models trained with Geti AI technology automatically generate exciting highlights, creating customized segments from the key moments of numerous competitions, and further attracting more users with personalized content. By accurately capturing and analyzing real-time match data, it not only provides a reference for referees to reduce human errors but also enhances the transparency and credibility of judgments, ensuring the fairness and impartiality of the competitions.
“Intel Geti AI”From Intel
Potential AI technologies involved :
1. Artificial Intelligence Generated Content: Automatically generating exciting highlights through models trained with Intel Geti AI technology.
2. Personalized Recommendations: Pushing customized segments based on user preferences.
3. Visual Content Processing: Precisely capturing and analyzing competition data.
4. Real-time Motion Analysis: In track and field, gymnastics, and other competitions, AI precisely analyzes the trajectory and force of athletes’ movements.
5. Real-time Location Monitoring: Monitoring the location and status of athletes in competitions such as canoeing, marathon, and cycling to assist referees in making accurate judgments.
Dataocean’s dataset of over 50,000 sports videos
Leveraging these cutting-edge AI technologies enhances the fairness of competitions and the viewing experience, offering a better interpretation of the Olympic spirit of peace, respect, friendship, and excellence. Behind these technologies lies the reliance on a vast amount of high-quality training data.
Dataocean AI’s Sport Videos with Highlights Dataset comprises over 50,000 sports action videos, totaling 51,618, with key moments in the actions labeled with start and end frames. The dataset covers a variety of sports including basketball, football, badminton, table tennis, skateboarding, street dance, gymnastics, track and field, high jump, long jump, golf, and dance. Each sport category contains approximately 1,000 to 3,000 videos, with each video featuring at least 1 to 3 valid actions, and the entire video length is within one minute.
