Journal of Multiscale Materials Informatics

A novel quantum circuit-based adaptive quantum convolutional neural network for image classification

Reza Pamungkas Putra Sukanli, Muhamad Akrom — Sat, 25 Jul 2026 00:00:00 +0000

This study explores Quantum Convolutional Neural Network (QCNN) starting from foundational quantum operations, such as the Rx gate for encoding MNIST image data into quantum states. We implemented quantum convolutional and pooling layers using one_unitary and two_unitary circuits, enabling effective feature extraction and dimensionality reduction while preserving critical information. Expressibility analysis revealed varying capabilities across different one_unitary circuits, with Rx, Ry, and Rz combinations demonstrating promising results akin to Haar random states. The proposed QCNN model exhibited robust performance metrics (accuracy: 95.98%, precision: 94.44%, recall: 96.59%, F1-score: 0.9551, AUC: 0.9604) in classification tasks, supported by efficient convergence during optimization. Future directions include expanding QCNN applications to handle more complex datasets and optimizing architectures to enhance quantum machine learning capabilities, particularly in image processing. This study underscores the potential of QCNNs in advancing quantum computing applications in neural network architectures.

Resource-Constrained Perception: Deploying Quantized Deep Learning Models on Microcontrollers for Real-Time Robotic Vision

Enjoy Bhodra, Md Nahiduzzaman Hridoy, M A Shahriar, Iffat Salim Toaha — Mon, 31 Aug 2026 00:00:00 +0000

Robotic vision enables machines to "see" and interpret the world around them. It can help with tasks such as obstacle avoidance, navigation, and object detection. This kind of work is well suited to modern deep learning but requires a lot of memory, power, and processing speed. Most of the small robots do not have a powerful computer. However, they have a low-cost microcontroller. This is a smaller computer. These devices are very resource-constrained. It becomes difficult to run regular deep learning models. This research work proposes to address the above issue with model quantization. When a model number is not a nice integer. It is quantized, i.e., rounded to a simpler integer with 8 bits. Such a reduction reduces memory consumption and increases the computing speed. This work is a comparison of two popular approaches. They are Post-Training Quantization (PTQ) and Quantization-Aware Training (QAT). Researchers construct a small and simple ConvNet that has only 28,069 parameters. It is examined on two popular microcontrollers: ARM Cortex M7 and ESP32. Both methods have been found to be effective. By reducing the use of the flash memory by approximately 71% and RAM by 60%, quantization cuts down memory usage. It doubles the processing speed and reduces power consumption by 35 to 40 percent. QAT retains 99% of the original model accuracy. The advantages of PTQ are that it is faster to set up, but with slightly less accuracy. The model performs well in real time on both platforms, with no assistance from the clouds. This work shows deep learning can run well on low-power devices. It helps make smart, low-cost robots more widely available.

Implementation of a Finite State Machine for Controlling NPC Customer Behavior in the 2D Game "Cooking Chaos"

Syakira Tasya, Diny Syarifah Sany — Thu, 03 Sep 2026 00:00:00 +0000

Non-Player Characters (NPC) are components in games that can be used to create interactions and gameplay dynamics. NPC behavior control is required to enable characters to respond appropriately to conditions occurring during gameplay. This study aims to implement a Finite State Machine (FSM) to control customer NPC behavior in the Unity-based 2D game Cooking Chaos. The implemented FSM consists of four states: Waiting, Angry, Chasing, and Attack. State transitions are influenced by the customer’s patience system and interaction conditions with the player. When the patience time expires, the customer changes behavior from waiting to becoming angry, chasing, and potentially attacking the player. The implementation was evaluated using Black Box Testing, FSM Transition Testing, and Multi-Customer Testing. The Black Box Testing results showed that all 12 functional scenarios were valid. FSM Transition Testing on seven scenarios, each tested 10 times, resulted in 70 successful trials out of 70. Multi-Customer Testing demonstrated that the order, patience, and FSM mechanisms operated independently with up to three active customers simultaneously. The results indicate that the implemented FSM consistently controls customer NPC behavior transitions under the defined test scenarios and integrates the service mechanism with direct gameplay consequences for the player.

Graph Neural Networks in Bioinformatics: From Biological Graph Construction to Interpretable and Scalable Precision Medicine

Seyed Enayatallah Alavi — Fri, 30 Oct 2026 00:00:00 +0000

Graph neural networks (GNNs) are increasingly used to integrate biological entities and relations across proteins, genes, drugs, molecular structures, and patients. This integrative narrative review critically examines how graph provenance, model architecture, evaluation design, explanation quality, and deployment constraints jointly determine the credibility of GNN findings in bioinformatics. In contrast to primarily architecture- and application-oriented surveys, we synthesize empirical case studies and propose a task-to-graph-to-validation reporting framework. We compare GCN, GAT, GraphSAGE, geometric/equivariant models, graph transformers, and graph–sequence hybrids; examine study-level benchmarks and baselines; and distinguish random, scaffold, protein-disjoint, temporal, and external validation. We also assess multi-omics fusion, graph-specific explanations, and computational scaling. Published examples show useful task-specific performance but do not justify universal superiority claims. We conclude that prospective utility depends on leakage-resistant evaluation, traceable graph construction, faithful and stable explanations, and independent biological or clinical verification.