Skip to main content
. 2024 Sep 3;24(17):5724. doi: 10.3390/s24175724

Table 3.

Performance of different unimodal models on the LFFD dataset under SIE protocol and comparison with the state of the art.

Angry (%) Happy (%) Neutral (%) Surprise (%) Avg (%) ± SD (%)
Model of [11] (SA images) 80.50 86 71.50 83.50 80 ± 9.03
Model_1 with AiF images 88.57 97.14 92.86 97.87 94.11 ± 4.08
Model_2 with Depth images 64.29 67.14 48.57 57.86 59.46 ± 7.17
model_3 with SA images 86.67 95 94.17 91.67 91.88 ± 3.25