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

SEED

Benchmarks

Task NameDataset NameSOTA ResultTrend
Multimodal UnderstandingSEED
Accuracy82.6
226
Emotion RecognitionSEED-IV Session 2 → Session 3
Accuracy100
198
Emotion RecognitionSEED-IV Session 1 → Session 3
Accuracy97.69
198
Emotion RecognitionSEED-IV Session 1 → Session 2
Accuracy100
134
Multimodal UnderstandingSEED-2-Plus
Accuracy81.1
125
Multimodal UnderstandingSEED-I Image
Accuracy0.822
75
EEG Emotion RecognitionSEED
Accuracy96.55
74
Emotion RecognitionSEED Session 1 → Session 2
Accuracy100
70
Multimodal UnderstandingSEED
SEED Score81.8
51
Emotion RecognitionSEED, SEED-IV, and SEED-V
Transfer Accuracy (SEED → SEED-IV)57.4
44
Text-to-SpeechSeed ZH
CER0.77
42
Text-to-SpeechSeed EN
WER1.23
41
Visual Question AnsweringSEED
Overall Score79.8
35
EEG emotion recognitionSEED Subject-independent
Accuracy90.92
34
EEG classificationSEED
AUPRC74.25
32
Text-to-SpeechSeed-zh (test)
CER0.78
32
Emotion RecognitionSEED-V (LOSO)
Accuracy62.77
32
Emotion RecognitionSEED (cross-subject cross-session)
Accuracy93.18
30
Text-to-SpeechSeed-en (test)
WER1.31
30
Multi-modal UnderstandingSEED-IMG
Accuracy78.46
29
Emotion RecognitionSEED IV V
Accuracy (SEED -> SEED-IV)56.29
28
Emotion RecognitionSEED VII
Balanced Accuracy33.56
28
EEG emotion recognitionSEED-IV (Subject-independent)
Accuracy79.3
28
Multimodal UnderstandingSEED-IMG
Accuracy71.9
27
Emotion RecognitionSEED cross-subject
Accuracy94.53
26
Showing 25 of 127 rows