Infertility has become a common problem in global health, and unsurprisingly, many couples need medical assistance to achieve reproduction. Many human behaviors can lead to infertility, which is none other than unhealthy sperm. The important thing is that assisted reproductive techniques require selecting healthy sperm. Hence, machine learning algorithms are presented as the subject of this research to effectively modernize and make accurate standards and decisions in classifying sperm. In this study, we developed a deep learning fusion architecture called SwinMobile that combines the Shifted Windows Vision Transformer (Swin) and MobileNetV3 into a unified feature space and classifies sperm from impurities in the SVIA Subset-C. Swin Transformer provides long-range feature extraction, while MobileNetV3 is responsible for extracting local features. We also explored incorporating an autoencoder into the architecture for an automatic noise-removing model. Our model was tested on SVIA, HuSHem, and SMIDS. Comparison to the state-of-the-art models was based on F1-score and accuracy. Our deep learning results accurately classified sperm and performed well in direct comparisons with previous approaches despite the datasets’ different characteristics. We compared the model from Xception on the SVIA dataset, the MC-HSH model on the HuSHem dataset, and Ilhan et al.’s model on the SMIDS dataset and the astonishing results given by our model. The proposed model, especially SwinMobile-AE, has strong classification capabilities that enable it to function with high classification results on three different datasets. We propose that our deep learning approach to sperm classification is suitable for modernizing the clinical world. Our work leverages the potential of artificial intelligence technologies to rival humans in terms of accuracy, reliability, and speed of analysis. The SwinMobile-AE method we provide can achieve better results than state-of-the-art, even for three different datasets. Our results were benchmarked by comparisons with three datasets, which included SVIA, HuSHem, and SMIDS, respectively (95.4% vs. 94.9%), (97.6% vs. 95.7%), and (91.7% vs. 90.9%). Thus, the proposed model can realize technological advances in classifying sperm morphology based on the evidential results with three different datasets, each having its characteristics related to data size, number of classes, and color space.

Penelitian ini bertujuan untuk mengembangkan trainer PID controller dan menilai kelayakan trainer PID Controller. Trainer yang dikembangkan menggunakan aplikasi PID controller yaitu pengendali dual brushless. Penelitian ini merupakan penelitian R&D yang berorientasi produk berupa trainer dan modul pendamping. Setelah produk mendapat validasi dari ahli melalui uji alfa dan uji beta maka produk dievaluasi pada implementasi sesungguhnya. Pengembangan media merujuk pada model Lee & Owens. Penelitian ini menggunakan model pengembangan yang meliputi analisis, desain, implementasi dan evaluasi, dimana letak evaluasi dan revisi terletak pada tiap tahapan. Evaluasi dilakukan oleh ahli materi, ahli media dan pengguna. Pengambilan data menggunakan angket yang memiliki rentang nilai 1 s.d. 4. Perolehan nilai angket selanjutnya dikonversi untuk mengetahui kategori kelayakan. Hasil penelitian menunjukkan nilai kelayakan trainer PID Controller adalah 3,33 dan termasuk kategori Sangat Baik. Karena trainer PID Controller memperoleh nilai kelayakan sangat baik maka trainer ini dapat digunakan dalam pembelajaran di kelas

Penelitian ini bertujuan menghasilkan trainer Internet of Things untuk mahasiswa jurusan pendidikan teknik Informatika dan menilai kelayakan trainer Internet of Things untuk mahasiswa jurusan pendidikan teknik Informatika. Pada trainer yang dikembangkan, aplikasi trainer Internet of Things digunakan untuk kendali jarak jauh menggunakan internet. Penelitian ini merupakan penelitian R&D karena hasil penelitian berorientasi pada produk yang berupa trainer dan modul pendamping. Penelitian ini menggunakan model pengembangan yang terdiri dari analisis, desain, implementasi dan evaluasi, dimana letak evaluasi dan revisi berada pada setiap tahapan. Pengambilan data menggunakan angket yang memiliki rentang nilai 1 s.d. 4. Perolehan nilai angket selanjutnya dikonversi untuk mengetahui kategori kelayakan. Hasil penelitian menunjukkan bahwa nilai kelayakan media trainer Internet of things secara umum memperoleh nilai 3,33 yang diintrepetasikan dalam kategori Sangat Baik dan diperoleh beberapa masukan yang telah diperbaiki. Nilai kelayakan yang dapat diintrepetasikan sangat baik pada trainer Internet of Things menunjukkan bahwa trainer ini dapat digunakan dalam pembelajaran dikelas.

Electronic noses (E-Noses) offer a rapid and non-destructive way to identify substances. However, accurately classifying complex sensor data from E-Noses remains a challenge. While effective at handling this data, deep learning suffers from high computational demands, hindering its use in portable devices for resource-constrained environments. Furthermore, there is an opportunity to better exploit the interplay between temporal, spatial, and spectral information within E-Nose data. This study addresses these limitations by proposing a novel feature extraction technique called Temporal-Spatial and Signal Descriptor (TSSD). TSSD leverages both time-domain and frequency-domain data to capture key odor signatures. It builds upon existing advancements to extract a more comprehensive feature space for accurate odor classification using machine-learning algorithms. We validate TSSD’s effectiveness against existing approaches using a public E-Nose dataset with 3, 5, and 10-fold cross-validation. Additionally, we compare its performance to 7 datasets with its state-of-the-art methods. Our approach demonstrated that it has the potential to improve E-nose classification accuracy and outperform the benchmark while enabling computation costs around 13ms to 14.4ms.

This letter focuses on developing practical and resource-efficient solutions for indoor positioning systems using Inertial Measurement Unit sensors (IMU) by introducing a compact and efficient model. The model, derived from the RoNIN architecture, features a lightweight model that is achieved by reducing the number of filters. A specific Knowledge Distillation (KD) method, Regression Confidence Forge (ReCoF) KD, is proposed and employed to address potential performance implications, enhancing the efficacy of the streamlined model. The smallest proposed model exhibits an 86% size reduction from RoNIN Resnet, leading to an 18.8% acceleration in inference time and 56% more power efficiency on the edge. Notably, the proposed model maintains high performance, as evidenced by its absolute trajectory error (ATE) and relative trajectory error (RTE).