Proposed KAN for deepfake identification on resource constrained environments.

<p>Proposed KAN for deepfake identification on resource constrained environments.</p>

محفوظ في:

التفاصيل البيبلوغرافية
المؤلف الرئيسي:	Nadeem Jabbar (21688844) (author)
مؤلفون آخرون:	Sohail Masood Bhatti (19105561) (author), Muhammad Rashid (2836613) (author), Arfan Jaffar (19105558) (author), Sheeraz Akram (17251060) (author)
منشور في:	2025
الموضوعات:	Cell Biology Molecular Biology Biotechnology Science Policy Space Science Biological Sciences not elsewhere classified Information Systems not elsewhere classified xlink "> deepfakes traditional detection methods experimental results show convolutional neural networks 21 million flops 11 million parameters time deepfake detection specifically sdg 9 faceforensics ++ dataset efficient deepfake classification kan achieves 95 deepfake classification sdg 16 df dataset work aligns untested capabilities study evaluates significantly less layer kolmogorov layer kan iot systems future research edge devices digital content balancing efficiency art cnns arnold network adversarial attacks 4 mb 200 nodes
الوسوم:	إضافة وسم لا توجد وسوم, كن أول من يضع وسما على هذه التسجيلة!

مواد مشابهة