Pith. sign in

This would allow us to explore in greater depth the differences in modalities if they are present

FUTURE WORK In the future we hope to apply our method to larger, more recent models supported by a corresponding increase in access to hardware

1 Pith paper cite this work. Polarity classification is still indexing.

1 Pith paper citing it
browse 1 citing papers

fields

cs.CV 1

years

2025 1

verdicts

REJECT 1

representative citing papers

Explaining Automatic Image Assessment

cs.CV · 2025-02-03 · reject · novelty 4.0

Training separate NIMA-style models on depth, saliency, and blur versions of AVA shows saliency carries the most signal among non-RGB modalities, while the standard 5.0 threshold inflates baselines to above 70 percent.

citing papers explorer

Showing 1 of 1 citing paper.

  • Explaining Automatic Image Assessment cs.CV · 2025-02-03 · reject · none · ref 6

    Training separate NIMA-style models on depth, saliency, and blur versions of AVA shows saliency carries the most signal among non-RGB modalities, while the standard 5.0 threshold inflates baselines to above 70 percent.