JOHOR BAHRU, Apr 19 — A dynamic one-day workshop titled “AI Workshop with TensorFlow” was successfully held at UTM Makerspace, co-organized by the Digital Signal and Image Processing (DSIP) Research Group, the Faculty of Electrical Engineering (FKE), the AIROST student society, and the Institute of Electrical and Electronics Engineers (IEEE) Signal Processing Society Malaysia Chapter. The event brought together over 40 tech enthusiasts from various universities for an immersive experience in machine learning. The workshop featured lectures from Associate Professor Dr. Zaid Omar, Associate Professor Dr. Usman Ullah Sheikh, and Tn. Hj. Muhammad Mun’im Ahmad Zabidi.
Participants were introduced to the fundamentals of machine learning, covering key topics such as neural network construction, computer vision, and model deployment. Through a combination of lectures, live coding sessions, and practical demonstrations, attendees gained firsthand experience in building and deploying Artificial Intelligence (AI) models.
Convolutional Neural Networks (CNN) are part of deep learning, a subset of machine learning within the broader field of artificial intelligence (AI). Unlike traditional machine learning methods, CNNs and other deep learning algorithms can process raw data without prior processing, which simplifies the task for the user. Users need only to determine the specific application and the appropriate dataset. To implement machine learning, various application programming interfaces (APIs) are available. For instance, scikit-learn is commonly used for traditional machine learning, while TensorFlow, which incorporates Keras, supports both machine learning and deep learning. TensorFlow enables users to create their own machine-learning models and also assists with inputting data, training the model, and deploying it via TensorFlow Serving.
Highlights of the program included an overview of computer vision and image processing, a step-by-step walkthrough of classification and CNN, building and training deep learning models using TensorFlow and Keras, implementing image classification for digit recognition, and deploying models with TensorFlow Lite on Google Colab.
“This workshop provides an excellent starting point for students new to deep learning, offering practical experience with TensorFlow and its applications. It is a valuable opportunity to enhance technical competence and engage with peers in the AI community,” said Najamuddin, a PhD candidate.
“Students will gain practical experience in developing and training machine learning models, which will strengthen their technical and research skills. It also exposes them to real-world AI-related applications, better preparing them for advanced academic projects or industry roles,” noted Habibullah, another PhD candidate.
The organizers aim to build on this momentum by offering more advanced training opportunities in the near future. Moreover, the conduct of this program aligns with Malaysia’s aspirations to lead in the advancement of AI technologies across various applications. This initiative supports the country’s goal of attaining high-income and developed nation status by 2028.
