{"ID":2845043,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2511.05479","arxiv_id":"2511.05479","title":"Quantization Effects of Artificial Neural Networks for Embedded Edge-Computing Applications","abstract":"This paper examines the use of Quantized Neural Networks (QNNs) for two resource-constrained scientific applications: automated calibration of semi-conductor quantum bits (qubits) and scientific particle detectors. We evaluate the trade-offs between Post-Training Quantization (PTQ), Quantization-Aware Training (QAT), and ultra-low-bit Binary Neural Networks (BNNs) with respect to latency and resource usage. Our results demonstrate that PTQ achieves a four-fold reduction in memory usage for U-shaped CNN (U-Net) architectures while maintaining or slightly enhancing segmentation accuracy (e.g. from 89% to 90% for a small U-Net with 447 parameters). For the training of non-differentiable custom BNNs , we propose a novel, hardware-constrained learning approach using Genetic Algorithms (GAs). We showcase a LUT-based BNN architecture suitable for direct conversion to VHDL via the HCL4BNN framework. This method achieves nanosecond-scale inference latencies (10-15 ns) without requiring specialized DSP or BRAM resources.","short_abstract":"This paper examines the use of Quantized Neural Networks (QNNs) for two resource-constrained scientific applications: automated calibration of semi-conductor quantum bits (qubits) and scientific particle detectors. We evaluate the trade-offs between Post-Training Quantization (PTQ), Quantization-Aware Training (QAT), a...","url_abs":"https://arxiv.org/abs/2511.05479","url_pdf":"https://arxiv.org/pdf/2511.05479v3","authors":"[\"Alperen Aksoy\",\"Ilja Bekman\",\"Vesselin Dimitrov\",\"Qader Dorosti\",\"Chimezie Eguzo\",\"Sarah Fleitmann\",\"Christian Grewing\",\"Fabian Hader\",\"Andre Zambanini\",\"Stefan van Waasen\"]","published":"2025-11-07T18:44:50Z","proceeding":"cs.NE","tasks":"[\"cs.NE\",\"physics.ins-det\"]","methods":"[\"Convolutional Neural Network\"]","has_code":false}