CT-DegradBench is a new benchmark dataset for CT degradation detection and severity estimation, with the SeSpeCT framework using semantic priors from vision-language models and frequency features to predict artifact types and levels without fine-tuning.
Per-Degradation Classification Performance Table 5 reports classification accuracy and F1-score for each degradation across single distortions (S1–S5) and mix- tures (M1–M5)
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
CT-DegradBench: A Physics-Informed Benchmark for CT Degradation Detection and Severity Estimation
CT-DegradBench is a new benchmark dataset for CT degradation detection and severity estimation, with the SeSpeCT framework using semantic priors from vision-language models and frequency features to predict artifact types and levels without fine-tuning.