Institutional Repository UIN Sunan Kalijaga YogyakartaOPTIMALISASI CNN RINGAN MENGGUNAKAN SPECAUGMENT DAN SHARPNESS-AWARE MINIMIZATION UNTUK MENINGKATKAN GENERALISASI PADA KLASIFIKASI SUARA BURUNGNIM.: 24206052003David SuharjantoauthorBird sound classification based on artificial intelligence is a promising approach for supporting automatic bird species identification. However, the generalization capability of deep learning models remains a challenge, particularly when confronted with noisy environments and out-of-distribution (OOD) data. This study aims to evaluate the effects of SpecAugment (Aug) and Sharpness-Aware Minimization (SAM) on the performance, robustness, generalization capability, and computational efficiency of lightweight convolutional neural network (CNN) models for bird sound classification. Experiments were conducted using the Birdsdata dataset as the primary dataset, UrbanSound8K for robustness evaluation under noisy conditions, and the Shorebird dataset as the OOD dataset. Three lightweight CNN architectures were evaluated: MobileNetV3-Small, EfficientNet-B0, and ConvNeXt-Atto. The evaluation included performance on clean data, robustness under noise at signal-to-noise ratio (SNR) levels ranging from -5 to 20 dB, generalization capability on OOD data, along with confusion matrix and Grad-CAM analyses, and inference efficiency on the Radxa ROCK 4 SE edge device. The results demonstrate that the combination of SpecAugment and SAM consistently improved the performance of all evaluated architectures compared with the baseline models, yielding F1-score improvements of 2.9-5.8 percentage points. ConvNeXt-Atto achieved the best performance with an accuracy of 96.7% and an F1-score of 96.1%, while also exhibiting superior robustness to noise and improved OOD detection capability based on the AUROC metric. These findings were supported by confusion matrix and Grad-CAM analyses. In edge deployment, ConvNeXt-Atto achieved the best trade-off between accuracy and efficiency, with 484.3 ms inference latency and 356.6 MB RAM usage. These findings demonstrate that SpecAugment and Sharpness-Aware Minimization improve the accuracy, robustness, and generalization of lightweight CNNs while maintaining computational efficiency.004 Pemrosesan Data, Ilmu Komputer, Teknik Informatika2026-07-21UIN SUNAN KALIJAGA YOGYAKARTA;FAKULTAS SAINS DAN TEKNOLOGIThesis

For work being deposited by its own author: In self-archiving this collection of files and associated bibliographic metadata, I grant Institutional Repository UIN Sunan Kalijaga Yogyakarta the right to store them and to make them permanently available publicly for free on-line. I declare that this material is my own intellectual property and I understand that Institutional Repository UIN Sunan Kalijaga Yogyakarta does not assume any responsibility if there is any breach of copyright in distributing these files or metadata. (All authors are urged to prominently assert their copyright on the title page of their work.)

For work being deposited by someone other than its author: I hereby declare that the collection of files and associated bibliographic metadata that I am archiving at Institutional Repository UIN Sunan Kalijaga Yogyakarta) is in the public domain. If this is not the case, I accept full responsibility for any breach of copyright that distributing these files or metadata may entail.

Clicking on the deposit button indicates your agreement to these terms.