Answers are grounded in this paper and cite their sources.
detail [15]. This limits the ability to train models 111 that can generalize well to such conditions. Foggy 112 Zurich [3] and Foggy Cityscapes [4] introduced 113 synthetic fog images by applying masks to original 114 images. Synthetic rain datasets [14] include 115 Rain1400, RainyCityscapes Rain100H, and 116 Rain12. Datasets like, BDD100K, ACDC [11], and 117 IDD [1], contain authentic images collected from 118 various adverse conditions [15], including fog, 119 rain, snow, and nighttime scenarios [5] [6]. The label set used in dataset preparation is identical to that in IDD [1] The hierarchy in the category of class labels adds a higher degree of complexity [5] to our dataset compared to existing datasets like Cityscapes [7], and even when compared to adverse weather datasets such as Foggy Cityscapes [8]
No takes yet. Share an insight, caveat, or question.
Mane et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: