Abstract The study's primary objective is the development of a framework that operates in real time and can determine the emotional state that most people in a group are experiencing on average. This research article suggests a basic method for recognising facial expressions by combining transfer learning with convolutional neural networks (CNNs) that have few parameters (TL). The suggested CNN architecture was jointly trained on the FER-2013, JAFFE, and CK+ datasets for real-time detection, which expanded the scope of what can be detected emotionally in terms of expressions of emotion. The model could notify when a person was happy, sad, surprised, scared, angry, disgusted, or neutral. Several methods were used to sort out how well the model worked. Experimental findings support the claim that the proposed approach is more effective than other research in terms of both precision and speed.
Alan : Mühendislik
Dergi Türü : Uluslararası
Benzer Makaleler | Yazar | # |
---|
Makale | Yazar | # |
---|