MoTIF adds temporal self-attention and automatic VLM-based concept discovery to concept bottleneck models for interpretable video classification, showing gains over prior global CBMs on benchmarks.
Invertible concept-based explanations for cnn models with non-negative concept activation vectors
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Concepts in Motion: Temporal Concept Bottleneck Model for Interpretable Video Classification
MoTIF adds temporal self-attention and automatic VLM-based concept discovery to concept bottleneck models for interpretable video classification, showing gains over prior global CBMs on benchmarks.