ConViS-Bench
A benchmark dedicated to evaluating concept-level similarity between videos, covering multiple concept dimensions such as activities, locations, and objects. Models are required to judge whether two videos are consistent with respect to a specified concept.
Timeline (last 90 days)
VICE was evaluated on two benchmarks, ConViS-Bench and StepDiff, and outperformed models that process two videos simultaneously.
ConViS-Bench is a benchmark dedicated to evaluating concept-level video similarity, covering multiple concept dimensions such as activity, location, and objects, and requiring models to judge whether two videos agree on a specified concept.
All Facts (2)
VICE was evaluated on two benchmarks, ConViS-Bench and StepDiff, and outperformed models that process two videos simultaneously.
50%UnverifiedConViS-Bench is a benchmark dedicated to evaluating concept-level video similarity, covering multiple concept dimensions such as activity, location, and objects, and requiring models to judge whether two videos agree on a specified concept.
50%