Object Hallucination Evaluation
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
94.27Random Score
51
Apr 7, 2026
54.6CHAIRs
51
Apr 7, 2026
90.13Accuracy
43
Jun 30, 2026
91.5Accuracy
43
Jun 30, 2026
93.87Accuracy
43
May 15, 2026
62CS Score
42
May 26, 2026
91.56Accuracy (POPE)
38
Apr 29, 2026
56.4CHAIR Scene Score
35
Mar 31, 2026
84.27Accuracy
34
May 15, 2026
90.81Accuracy
32
Jun 26, 2026
86.07Accuracy
31
Jun 30, 2026
57CHAIRS
30
Jun 1, 2026
84.72Average Accuracy
29
Apr 1, 2026
32.4CHAIRs
28
Apr 23, 2026
89.2Accuracy
27
Jun 30, 2026
52.4CHAIR_S
27
Feb 26, 2026
88.7Random F1-score
26
May 25, 2026
86.15Average F1
24
Jun 30, 2026
43.7CHAIRS Score
24
May 28, 2026
92.23Accuracy
24
Jun 1, 2026
73.5Accuracy
24
Apr 23, 2026
80.05Accuracy
24
Mar 12, 2026
82.23Accuracy
24
Mar 12, 2026
86.79Accuracy
24
Mar 12, 2026
89.1POPE Score
22
Jun 24, 2026