Yazar "Özcan, Ahmet Remzi" seçeneğine göre listele

Listeleniyor 1 - 7 / 7
  • Küçük Resim Yok
    Öğe
    A Ground-Truth-Free Framework for Validating Emotions in Generative AI Speech Synthesis
    (Institute of Electrical and Electronics Engineers Inc., 2026) Özcan, Ahmet Remzi
    Evaluating emotional expressivity in synthetic speech is challenging due to the absence of ground-truth affective labels and the reliance on costly human perceptual studies. This paper introduces a prototype-based framework that integrates affect-specialized Emotion2Vec embeddings with general-purpose acoustic and linguistic representations from WavLM to enable scalable and system-agnostic evaluation. Embeddings are projected into a shared latent space where each emotion category is represented by a learnable prototype, supporting both categorical classification and a continuous similarity-based metric, the Emotion Adherence Score (EAS). While categorical performance varied across systems, EAS remained consistently high, highlighting its robustness in capturing graded affective fidelity. On a 1,400-utterance corpus spanning four heterogeneous TTS systems, the proposed method achieved substantial improvements over a strong embedding baseline, increasing accuracy from 51.43% to 77.50% and macro-F1 from 0.5109 to 0.7736. Human ratings further supported EAS, showing a moderate positive correlation with human judgments. Overall, the proposed framework provides a principled and scalable approach for benchmarking emotional expressivity in TTS, bridging categorical and continuous perspectives and reducing reliance on ground-truth labels and large-scale listening tests. © 2013 IEEE.
  • Küçük Resim Yok
    Öğe
    Deep Learning-Based Turkish License Plate Recognition System on Low-Power Microcontroller Systems
    (Institute of Electrical and Electronics Engineers Inc., 2024) Gorgulu, Emre; Özcan, Ahmet Remzi
    Automatic License Plate Recognition (ALPR) systems play a crucial role in intelligent transportation systems, traffic management, and security. However, ensuring that these systems operate efficiently on embedded devices with low power consumption presents a significant challenge. In this study, an energy-efficient ALPR system has been designed using the RISC-V-based Kendryte K210 microcontroller. The proposed system adopts a two-stage deep learning-based architecture consisting of YOLOv2 and LPRNet models for license plate detection and recognition. The YOLOv2 model achieves high accuracy in the license plate detection process, while the LPRNet model performs character recognition on the detected license plates. This deep learning-based approach offers an ideal solution for portable and embedded systems with low power consumption. The training of the models leveraged various open-access datasets, aiming to enhance the model's ability to adapt to different conditions. The findings of the study demonstrate that the developed system achieves high accuracy rates in both license plate detection and recognition processes. The performance observed in the license plate detection stage indicates that the system can reliably operate under different conditions. However, the high inference time observed in the license plate recognition process suggests that the model needs to be optimized for real-time applications. In future work, optimization and quantization techniques in deep learning models are planned to enhance the speed and efficiency of the license plate recognition process. Additionally, the applicability of different text recognition models for the license plate recognition problem in embedded systems will be explored. © 2024 IEEE.
  • Küçük Resim Yok
    Öğe
    Epileptic Seizure Prediction with Recurrent Convolutional Neural Networks
    (Ieee, 2017) Özcan, Ahmet Remzi; Erturk, Sarp
    In this paper, the use of recurrent convolutional neural networks for predicting epileptic seizures is proposed. Effective methods for predicting epileptic seizures need to be developed for the design of diagnostic and therapeutic techniques that will prevent or mitigate epileptic seizures. Studies show that epileptic seizures appear as a consequence of temporal and spatially developed processes in epileptic networks. In many studies using different linear and nonlinear methods of measurement, the result is that the measurements are differentiated before the epileptic seizure takes place. In this study, the features extracted by different methods from the multi-channel EEG signals are transformed into multi-spectral image series by projecting depending on the placement of the electrodes. Recurrent convolutional neural networks are trained with the obtained multi-spectral image sequences to reveal spatial and temporal correlations in multi-channel EEG signals before the epileptic seizure.
  • Küçük Resim Yok
    Öğe
    Java Card Tabanlı Platform Bağımsız Mobil İmzalama Sistemi
    (2025) Yıldız, Omer; Özcan, Ahmet Remzi
    Kripto para kullanıcıları için özel anahtarların güvenli biçimde saklanması, dijital varlık yönetiminin temel gereksinimidir. Ancak mevcut donanım cüzdanları (Ledger, Trezor vb.), yüksek maliyet, sınırlı mobil uyumluluk, platform bağımlılığı ve kaynak kodlarının kapalı olması nedeniyle yaygın kullanım ve entegrasyon açısından sınırlılıklar taşımaktadır. Özellikle iOS cihazlarla donanımsal entegrasyon eksikliği, kullanıcı deneyimini olumsuz etkilemektedir.Bu çalışmada, Java Card teknolojisi ve NFC üzerinden NDEF standardı kullanılarak geliştirilen, platformdan bağımsız ve uygulama gerektirmeyen bir dijital imzalama sistemi sunulmaktadır. Sistem, Java Card üzerinde izole çalışan iki applet’ten oluşmakta; biri NDEF mesajlarını ayrıştırmakta, diğeri ise kriptografik işlemleri donanım düzeyinde gerçekleştirmektedir. Android ve iOS cihazlarla uyumlu geliştirilen React Native mobil istemci üzerinden yapılan testlerde, sistemin kararlı iletişim, yüksek performans ve kullanıcı dostu etkileşim sağladığı gözlemlenmiştir.
  • Küçük Resim Yok
    Öğe
    Low-Cost Android Based Tele-Monitoring System for Body Temperature Measurement
    (Bursa Teknik Üniversitesi, 2021) Özcan, Ahmet Remzi; Mert, Ahmet
    The increasing number of pandemic issues require to pay attention to health conditions and social distance. The explicit sign of COVID-19 is body fever. It is a simple and affordable detection method when compared to other blood tests. However, it is required to be physically close to a visitor to measure body temperature. For this reason, we have designed and developed a low-cost microprocessor measurement system with an infrared non-contact temperature sensor, and a Bluetooth for sending to long distance. Android application has been developed to read and set the alarm function using a smart telephone or a tablet far from the visitors. With the help of these circuit and application designs, body temperatures can be checked from long distance considering the pandemic situations. The printed circuit of the microcontroller, Bluetooth, sensor, and light-dependent resistor (LDR) triggering are manufactured, and the software of the controller and the application are integrated and tested successfully. 
  • Küçük Resim Yok
    Öğe
    Optimisation of pedestrian detection system using FPGA-CPU hybrid implementation for vehicle industry
    (Inderscience Enterprises Ltd, 2019) Özcan, Ahmet Remzi; Taysanoglu, Vedat
    Improved image processing and developing technologies are rapidly expanding the application areas of image processing systems. In recent years, pedestrian detection systems have become one of the major safety technologies used in the automotive industry. This paper presents an optimised real-time pedestrian detection system using an FPGA-CPU based hybrid design. The histograms of oriented gradients (HOG) algorithm, which is extensively used for feature extraction in pedestrian detection applications, was implemented on a low-end FPGA. In the study, the original HOG descriptors are designed in low complexity without sacrificing performance. The obtained features were classified on a low-power single board computer with support vector machine (SVM). Tests with the INRIA pedestrian database show that the proposed model has high potential for use as a real-time low-cost pedestrian detection system in practice.
  • Küçük Resim Yok
    Öğe
    Seizure Prediction in Scalp EEG Using 3D Convolutional Neural Networks With an Image-Based Approach
    (Ieee-Inst Electrical Electronics Engineers Inc, 2019) Özcan, Ahmet Remzi; Erturk, Sarp
    Epileptic seizures occur as a result of a process that develops over time and space in epileptic networks. In this study, we aim at developing a generalizable method for patient-specific seizure prediction by evaluating the spatio-temporal correlation in the features obtained from multichannel EEG signals. Spectral band power, statistical moment and Hjorth parameters are used to reveal the frequency and time domain features of the EEG signals. The features are given as input to a convolutional neural network (CNN) by transforming into a sequence of multi-color images according to the topology of the EEG channels. The multi-frame 3D CNN model is proposed to evaluate the temporal and spatial correlation in training data collectively. The proposed 3D CNN model achieves a sensitivity of 85.7%, a false prediction rate of 0.096/h, and a proportion of time-in-warning of 10.5%, in the tests performed with 16 patients from the CHB-MIT scalp EEG dataset. The results show that the superiority of the proposed method to a Poisson based random predictor was statistically significant for 93.7% of the patients, at significance level of 0.05. Our experiments with various timing constraints show that epileptic stage lengths are an important factor affecting seizure performance. We present a subject-specific seizure prediction method that is robust for unbalanced data and can be generalized to any scalp EEG dataset without the need for subject-specific engineering.