[KongchangAI]
BenchmarkRTTS数据集

RTTS

The real task test set in the RESIDE benchmark, consisting of naturally captured hazy images that lack pixel-level ground truth. It is commonly used to evaluate, with the help of no-reference metrics, the generalization ability of dehazing models on real haze.

Timeline (last 90 days)

Nov 6

The paper conducted experiments on paired CDD-11 fog images and real RTTS scenes.

Unverified50%
Nov 6

RTTS comes from the real-world task test set of the RESIDE benchmark; its images are naturally captured foggy scenes lacking pixel-level ground truth, so they can only be evaluated with no-reference metrics.

Unverified50%

All Facts (2)

Source Articles