Hybrid Transformer-Sequencer Approach for Age and Gender Classification from In-Wild Facial Images

Proposing a hybrid model combining self-attention and BiLSTM approaches significantly improves age and gender classification accuracy.