A Harvard‑led research group has introduced BRIDGE, a novel system that translates footage from able‑bodied basketball into realistic wheelchair basketball sequences, aiming to close the analytical gap that has long limited parasports training.
The work was recognized with a Best Paper Award at the ACM Conference on Human Factors in Computing Systems (CHI), highlighting its contribution to making video analysis tools more accessible to para‑athletes and coaches. The project was spearheaded by senior author Hanspeter Pfister alongside co‑lead authors Chunggi Lee and Hayato Saiki.
Technical Innovation Behind the Transformation
BRIDGE employs a reconstruction pipeline that extracts players and the ball from broadcast video, then uses an embodiment‑aware visualization framework to decompose head, trunk and wheelchair base orientations and remap them for wheelchair sport contexts.
In experiments with 20 participants, the system produced more natural player postures and clearer tactical cues than conventional non‑disabled video, according to the researchers.
Broader Implications and Future Directions
The team envisions extending the concept of embodiment transformation to other adaptive sports, integrating augmented and virtual reality with artificial intelligence to further personalize training resources.
Funding for the project came from the National Science Foundation under grant No. CRCNS‑2309041 and the Harvard Data Science Initiative Trust in Science Fund Award, underscoring institutional support for inclusive technology development.