Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Sign Language Translation Efficiency on Architectural Efficiency Comparison
Loading...
6.7
Video Encoder Total Params (M)
Geo-Sign
5.572
13.186
20.8
28.414
Jun 26, 2026
Video Encoder Total Params (M)
Video Encoder Trainable Params (M)
Language Model Total Params (M)
Language Model Trainable Params (M)
Full Model Total Params (M)
Full Model Trainable Params (M)
Full Model Forward Latency (ms)
Updated 26d ago
Evaluation Results
Method
Method
Links
Video Encoder Total Params (M)
Video Encoder Trainable Params (M)
Language Model Total Params (M)
Language Model Trainable Params (M)
Full Model Total Params (M)
Full Model Trainable Params (M)
Full Model Forward Latency (ms)
Geo-Sign
Input Modality=Skeleto...
2026.06
6.7
6.7
582.4
582.4
589.1
589.1
2,550
Uni-Sign
Input Modality=Skeleto...
2026.06
9.7
9.7
582.4
582.4
592.1
592.1
416
C2RL
Input Modality=RGB-bas...
2026.06
11.7
11.7
680
680
691.7
691.7
-
SIGNET
Input Modality=Skeleto...
2026.06
17.4
1.5
582.4
582.4
599.8
583.9
320
Sign2GPT
Input Modality=RGB-bas...
2026.06
34.9
12.8
1,736.7
3.8
1,771.6
16.7
-
Feedback
Search any
task
Search any
task
