dinov2_Liveness_detection_v2.2.2

This model is a fine-tuned version of nguyenkhoa/dinov2_Liveness_detection_v2.2.1 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.0219
• Accuracy: 0.9934
• F1: 0.9934
• Recall: 0.9934
• Precision: 0.9934

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 5e-05
• train_batch_size: 768
• eval_batch_size: 8
• seed: 42
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 5
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Recall	Precision
0.069	0.3062	64	0.0579	0.9788	0.9790	0.9788	0.9790
0.0482	0.6124	128	0.0518	0.9814	0.9815	0.9814	0.9815
0.0404	0.9187	192	0.0325	0.9882	0.9882	0.9882	0.9882
0.0325	1.2249	256	0.0407	0.9855	0.9855	0.9855	0.9855
0.0322	1.5311	320	0.0281	0.9901	0.9901	0.9901	0.9902
0.0283	1.8373	384	0.0347	0.9884	0.9884	0.9884	0.9885
0.0256	2.1435	448	0.0271	0.9907	0.9907	0.9907	0.9907
0.0207	2.4498	512	0.0359	0.9874	0.9874	0.9874	0.9874
0.0192	2.7560	576	0.0253	0.9917	0.9916	0.9917	0.9917
0.017	3.0622	640	0.0272	0.9908	0.9908	0.9908	0.9908
0.0134	3.3684	704	0.0255	0.9916	0.9916	0.9916	0.9915
0.0132	3.6746	768	0.0232	0.9925	0.9925	0.9925	0.9925
0.0114	3.9809	832	0.0260	0.9919	0.9919	0.9919	0.9918
0.0074	4.2871	896	0.0242	0.9927	0.9927	0.9927	0.9927
0.0079	4.5933	960	0.0219	0.9931	0.9931	0.9931	0.9932
0.0072	4.8995	1024	0.0219	0.9934	0.9934	0.9934	0.9934

Framework versions

• Transformers 4.47.0
• Pytorch 2.5.1+cu121
• Datasets 3.2.0
• Tokenizers 0.21.0

Evaluate results

• APCER: 0.1804
• BPCER: 0.0123
• ACER: 0.0963
• Accuracy: 0.8706
• F1: 0.8982
• Recall: 0.9877
• Precision: 0.7046