Sora Generates Videos with Stunning Geometrical Consistency
This work addresses a problem for researchers and practitioners in video generation by providing a quantitative evaluation method, though it is incremental as it builds on existing 3D reconstruction techniques.
The paper tackles the lack of metrics to evaluate Sora's video generation fidelity to real-world physics by introducing a benchmark that assesses video quality based on adherence to physics principles, using 3D reconstruction accuracy as a proxy.
The recently developed Sora model [1] has exhibited remarkable capabilities in video generation, sparking intense discussions regarding its ability to simulate real-world phenomena. Despite its growing popularity, there is a lack of established metrics to evaluate its fidelity to real-world physics quantitatively. In this paper, we introduce a new benchmark that assesses the quality of the generated videos based on their adherence to real-world physics principles. We employ a method that transforms the generated videos into 3D models, leveraging the premise that the accuracy of 3D reconstruction is heavily contingent on the video quality. From the perspective of 3D reconstruction, we use the fidelity of the geometric constraints satisfied by the constructed 3D models as a proxy to gauge the extent to which the generated videos conform to real-world physics rules. Project page: https://sora-geometrical-consistency.github.io/