Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Object Hallucination Evaluation on POPE Random offline

73.6F1 Score

Beam Search

52.259257.799663.3468.8804Jan 11, 2025
Updated 1mo ago

Evaluation Results

MethodLinks
2025.01
73.678.38
2025.01
72.9878.01
2025.01
72.9877.82
2025.01
72.8477.93
2025.01
72.3378.57
2025.01
72.1677.08
2025.01
71.7477.19
2025.01
71.0776.69
2025.01
70.4576.51
2025.01
70.2276.33
2025.01
70.1676.21
2025.01
69.8775.83
2025.01
69.3175.72
2025.01
69.1675.74
2025.01
69.0475.62
2025.01
65.3373.56
2025.01
62.1372.03
2025.01
62.13-
2025.01
61.6371.47
2025.01
61.63-
2025.01
60.4670.76
2025.01
60.46-
2025.01
60.3871.32
2025.01
60.38-
2025.01
60.0871.1
2025.01
60.08-
2025.01
60.0571.08
2025.01
60.05-
2025.01
59.871.02
2025.01
59.8-
2025.01
59.7370.82
2025.01
59.73-
2025.01
59.3670.73
2025.01
59.36-
2025.01
58.7570.55
2025.01
58.75-
2025.01
58.6370.49
2025.01
58.63-
2025.01
58.2868.78
2025.01
58.28-
2025.01
58.2770.38
2025.01
58.27-
2025.01
58.168.61
2025.01
58.1-
2025.01
57.469.67
2025.01
57.4-
2025.01
57.3669.84
2025.01
57.2469.77
2025.01
57.2169.61
2025.01
57.21-
2025.01
57.0569.32
2025.01
56.8669.13
2025.01
56.5569.14
2025.01
56.5369.05
2025.01
56.3169.02
2025.01
56.2969.66
2025.01
56.29-
2025.01
55.8269.34
2025.01
55.82-
2025.01
55.7868.9
2025.01
55.769.08
2025.01
55.7-
2025.01
55.3168.79
2025.01
55.31-
2025.01
54.8768.55
2025.01
54.87-
2025.01
53.9768.26
2025.01
53.97-
2025.01
53.9368.18
2025.01
53.93-
2025.01
53.9168.12
2025.01
53.91-
2025.01
53.8668.09
2025.01
53.86-
2025.01
53.8468.05
2025.01
53.84-
2025.01
53.8368.08
2025.01
53.83-
2025.01
53.7168.02
2025.01
53.71-
2025.01
53.2667.96
2025.01
53.26-
2025.01
53.0867.73
2025.01
53.08-