Academic Journal

Minimizing Multiplication of Kernel Computation in Convolutional Neural Networks Using Strassen Algorithm

Λεπτομέρειες βιβλιογραφικής εγγραφής
Τίτλος: Minimizing Multiplication of Kernel Computation in Convolutional Neural Networks Using Strassen Algorithm
Συγγραφείς: Rifqie, Dary Mochamad, Surianto, Dewi Fatmarani, Jayanegara, Sudarmanto, Fajar B, Muhammad, Fakhri, M. Miftach
Πηγή: Jurnal MediaTIK; Volume 6 Issue 2, Mei (2023); 52-54 ; 2715-5919 ; 2656-1247 ; 10.59562/mediatik.v6i2
Στοιχεία εκδότη: Jurusan Teknik Informatika dan Komputer
Έτος έκδοσης: 2024
Θεματικοί όροι: Convolutional Neural Network, Strassen Algorithm, Matrix Multiplication, Kernal Computation
Περιγραφή: Convolution neural networks (CNN) have been widely applied for the computer vision task. However, the success of CNN is limited by the computational complexity of the network, so it is difficult for the model to run the inference process in real time. In this paper, we apply Strassen matrix multiplication to reduce multiplications in convolution operations in CNN, in order to get faster execution for CNN. First, we transform the convolution operation into a matrix multiplication operation using the Toeplitz mapping method, then after that, we apply the Strassen method to these matrices. In the end, we compare the number of arithmetic operations (multiplication and addition) in the convolutional layer using Strassen and the standard algorithm. We apply this algorithm implementation in convolution layers 1 and 3 in LeNet-5 Architecture.
Τύπος εγγράφου: article in journal/newspaper
Περιγραφή αρχείου: application/pdf
Γλώσσα: English
Relation: http://journal.unm.ac.id/index.php/MediaTIK/article/view/1397/865; http://journal.unm.ac.id/index.php/MediaTIK/article/view/1397
DOI: 10.59562/mediatik.v6i2.1397
Διαθεσιμότητα: http://journal.unm.ac.id/index.php/MediaTIK/article/view/1397
https://doi.org/10.59562/mediatik.v6i2.1397
Rights: https://creativecommons.org/licenses/by-sa/4.0
Αριθμός Καταχώρησης: edsbas.FDA93BE6
Βάση Δεδομένων: BASE