Image Captioning
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
Image Captioning Dataset
82.9CLIP Score (RN-50)
30
Image Captioning
39.7BLEU-4
25
Image Captioning Hard Criterion Claude-3.5
7ASR
8
Image Captioning Hard Criterion GPT-o1
53ASR
8
Image Captioning Hard Criterion GPT-4o
74ASR
8
Image Captioning Target: Claude-4.5, Soft Criterion
8ASR
8
Image Captioning Target Gemini-2.5 Soft Criterion
79Accuracy Score Rate (ASR)
8
Image Captioning Target: GPT-5 Soft Criterion
56ASR
8
Image Captioning Target: GPT-4o, Soft Criterion
81ASR
8