Ahmet Emir Dirik
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Ahmet Emir Dirik has a total of 21 co-authors. Here are the top co-authors by number of shared publications:
Nasir Memon
7 shared publications- Ahmet Karaküçük
Department of Electricity and Energy, Bursa Uludağ University, Gemlik 16600, Bursa, Türkiye
5 shared publications - Saffet Vatansever
Bursa Teknik Üniversitesi
5 shared publications Husrev Taha Sencar
New York University Abu Dhabi, Abu Dhabi, United Arab Emirates
4 shared publications- Nasir Memon
Tandon School of Engineering, New York University, Brooklyn, NY, USA
3 shared publications Saffet Vatansever
Department of Mechatronics Engineering, Bursa Technical University, Bursa, Turkey
3 shared publications
Latest Publications
- Open Access
- 22 May 2026
- Published by PeerJ in PeerJ Computer Science
- Vol. 12, e3893
Assessing the visual realism of deepfake videos: an IQA-based approach
As deepfake videos become increasingly realistic, the ability of detection models to generalize across manipulation techniques with varying degrees of visual realism has become critical. However, many existing deepfake datasets contain visually low-realism samples that can hinder effective model training and result in misleading assessments of detection performance. While prior work has focused on dataset-level assessments of visual realism, systematic individual video-level research has received limited attention. To address this gap, this study proposes a video-level visual realism assessment method that quantifies distributional differences in no-reference image quality (NR-IQA) scores between pristine and manipulated videos using Cohen’s d as an effect-size measure, yielding a per-video score referred to as the Video Fidelity Score (VFS). Experiments conducted on the FaceForensics++, FaceShifter, and Celeb-DF datasets characterize realism variations both within individual datasets and across datasets under a consistent experimental setup. The resulting VFS-based rankings exhibit trends that are consistent with Fréchet Inception Distance (FID) measurements, supporting the reliability of the proposed realism stratification. In addition, a dedicated human evaluation study confirms that VFS aligns well with human visual perception when distinguishing between visually high- and low-realism videos. Overall, the proposed approach offers a practical tool for realism-aware dataset curation, with potential implications for both deepfake generation and detection research.
- 22 May 2026
- Published by PeerJ in PeerJ Computer Science
- Vol. 12, e3893
Open AccessAssessing the visual realism of deepfake videos: an IQA-based approach
As deepfake videos become increasingly realistic, the ability of detection models to generalize across manipulation techniques with varying degrees of visual realism has become critical. However, many existing deepfake datasets contain visually low-realism samples that can hinder effective model training and result in misleading assessments of detection performance. While prior work has focused on dataset-level assessments of visual realism, systematic individual video-level research has received limited attention. To address this gap, this study proposes a video-level visual realism assessment method that quantifies distributional differences in no-reference image quality (NR-IQA) scores between pristine and manipulated videos using Cohen’s d as an effect-size measure, yielding a per-video score referred to as the Video Fidelity Score (VFS). Experiments conducted on the FaceForensics++, FaceShifter, and Celeb-DF datasets characterize realism variations both within individual datasets and across datasets under a consistent experimental setup. The resulting VFS-based rankings exhibit trends that are consistent with Fréchet Inception Distance (FID) measurements, supporting the reliability of the proposed realism stratification. In addition, a dedicated human evaluation study confirms that VFS aligns well with human visual perception when distinguishing between visually high- and low-realism videos. Overall, the proposed approach offers a practical tool for realism-aware dataset curation, with potential implications for both deepfake generation and detection research.
- 10 April 2023
- Published by Omer Halisdemir Universitesi in Ömer Halisdemir Üniversitesi Mühendislik Bilimleri Dergisi
A Study on PRNU-based Source Device Identification for Digital Images and Videos of Different Resolutions
Her bir kamera sensörüne has benzersiz bir gürültü bileşeni olan PRNU (Photo Response Non-Uniformity), sayısal imge ve videoların adli analizi kapsamında ihtiyaç duyulan önemli araçlardandır. PRNU’nun en yaygın uygulama alanı olan kaynak kamera tespiti, aynı marka ve model kameraların bile PRNU karakteristiğinin birbirinden farklı oluşu ve bu örüntünün çekilen her bir resim karesi üzerine bir kamera parmak izi gibi istemsiz eklenmesi esasına dayanmaktadır. Bir test dosyasından (imge ya da video) kestirimi sağlanan PRNU sensör gürültüsü ile dosyanın kaynağı olduğu düşünülen kameraya ait referans PRNU (sabit içerikli düz duvar ya da gökyüzü görüntülerinden yüksek doğrulukta elde edilen PRNU örüntüsü) arasındaki benzerliğe göre bu kameranın test videosunun kaynağı olup olamayacağı belirlenebilir. Sayısal video çerçevelerinin imgelere göre düşük kalitede kodlanması, videolardan kestirilen PRNU sensör gürültüsünün doğruluğunu, dolayısıyla da benzerlik analizini etkilemektedir. Bu bağlamda, videolarda PRNU tabanlı kaynak kamera tespitinde, referans PRNU’nun videolardansa imgelerden elde edilmesi performans etkinliği için önemli bir hamledir. Ancak, imge ve videolar aynı kaynak kamera ile çekilmiş olsalar dahi farklı en boy oranında ve/veya çözünürlükte kaydedilmektedirler. Bu sebeple, imgelerden elde edilen PRNU izinin, sorgu videosuna ait PRNU sensör gürültüsü ile aynı foto-alıcı hücrelere karşılık gelecek forma dönüştürülmesi gerekmektedir. Bu çalışmada, bu dönüşümü sağlayan ölçekleme ve kırpma parametrelerini hassas bir şekilde hesaplayabilen bir yöntem önerilmiştir.
- 10 April 2023
- Published by Omer Halisdemir Universitesi in Ömer Halisdemir Üniversitesi Mühendislik Bilimleri Dergisi
A Study on PRNU-based Source Device Identification for Digital Images and Videos of Different Resolutions
Her bir kamera sensörüne has benzersiz bir gürültü bileşeni olan PRNU (Photo Response Non-Uniformity), sayısal imge ve videoların adli analizi kapsamında ihtiyaç duyulan önemli araçlardandır. PRNU’nun en yaygın uygulama alanı olan kaynak kamera tespiti, aynı marka ve model kameraların bile PRNU karakteristiğinin birbirinden farklı oluşu ve bu örüntünün çekilen her bir resim karesi üzerine bir kamera parmak izi gibi istemsiz eklenmesi esasına dayanmaktadır. Bir test dosyasından (imge ya da video) kestirimi sağlanan PRNU sensör gürültüsü ile dosyanın kaynağı olduğu düşünülen kameraya ait referans PRNU (sabit içerikli düz duvar ya da gökyüzü görüntülerinden yüksek doğrulukta elde edilen PRNU örüntüsü) arasındaki benzerliğe göre bu kameranın test videosunun kaynağı olup olamayacağı belirlenebilir. Sayısal video çerçevelerinin imgelere göre düşük kalitede kodlanması, videolardan kestirilen PRNU sensör gürültüsünün doğruluğunu, dolayısıyla da benzerlik analizini etkilemektedir. Bu bağlamda, videolarda PRNU tabanlı kaynak kamera tespitinde, referans PRNU’nun videolardansa imgelerden elde edilmesi performans etkinliği için önemli bir hamledir. Ancak, imge ve videolar aynı kaynak kamera ile çekilmiş olsalar dahi farklı en boy oranında ve/veya çözünürlükte kaydedilmektedirler. Bu sebeple, imgelerden elde edilen PRNU izinin, sorgu videosuna ait PRNU sensör gürültüsü ile aynı foto-alıcı hücrelere karşılık gelecek forma dönüştürülmesi gerekmektedir. Bu çalışmada, bu dönüşümü sağlayan ölçekleme ve kırpma parametrelerini hassas bir şekilde hesaplayabilen bir yöntem önerilmiştir.
- Open Access
A One-Dimensional Convolutional Neural Network-Based Method for Diagnosis of Tooth Root Cracks in Asymmetric Spur Gear Pairs
Gears are fundamental components used to transmit power and motion in modern industry. Their health condition monitoring is crucial to ensure reliable operations, prevent unscheduled shutdowns, and minimize human casualties. From this standpoint, the present study proposed a one-dimensional convolutional neural network (1-D CNN) model to diagnose tooth root cracks for standard and asymmetric involute spur gears. A 6-degrees-of-freedom dynamic model of a one-stage spur gear transmission was established to achieve this end and simulate vibration responses of healthy and cracked (25%–50%–75%–100%) standard (20°/20°) and asymmetric (20°/25° and 20°/30°) spur gear pairs. Three levels of signal-to-noise ratios were added to the vibration data to complicate the early fault diagnosis task. The primary consideration of the present study is to investigate the asymmetric gears’ dynamic characteristics and whether tooth asymmetry would yield an advantage in detecting tooth cracks easier to add to the improvements it affords in terms of impact resistance, bending strength, and fatigue life. The findings indicated that the developed 1-D CNN model’s classification accuracy could be improved by up to 12.8% by using an asymmetric (20°/30°) tooth profile instead of a standard (20°/20°) design.
Open AccessA One-Dimensional Convolutional Neural Network-Based Method for Diagnosis of Tooth Root Cracks in Asymmetric Spur Gear Pairs
Gears are fundamental components used to transmit power and motion in modern industry. Their health condition monitoring is crucial to ensure reliable operations, prevent unscheduled shutdowns, and minimize human casualties. From this standpoint, the present study proposed a one-dimensional convolutional neural network (1-D CNN) model to diagnose tooth root cracks for standard and asymmetric involute spur gears. A 6-degrees-of-freedom dynamic model of a one-stage spur gear transmission was established to achieve this end and simulate vibration responses of healthy and cracked (25%–50%–75%–100%) standard (20°/20°) and asymmetric (20°/25° and 20°/30°) spur gear pairs. Three levels of signal-to-noise ratios were added to the vibration data to complicate the early fault diagnosis task. The primary consideration of the present study is to investigate the asymmetric gears’ dynamic characteristics and whether tooth asymmetry would yield an advantage in detecting tooth cracks easier to add to the improvements it affords in terms of impact resistance, bending strength, and fatigue life. The findings indicated that the developed 1-D CNN model’s classification accuracy could be improved by up to 12.8% by using an asymmetric (20°/30°) tooth profile instead of a standard (20°/20°) design.
- 4 November 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Information Forensics and Security
- Vol. 18 (15566013) , 248-260
The Effect of Inverse Square Law of Light on ENF in Videos Exposed by Rolling Shutter
Due to a constant imbalance between demand and supply of power, ENF (Electric Network Frequency) fluctuates around a nominal value of 50 or 60 Hz. These variations in ENF cause the luminance intensity of a mains-powered light source, having no AC/DC converter inside, also to fluctuate. As a result, a video of a scene illuminated by a mains-powered light source can be used to estimate these fluctuations. As a consequence, the ENF signal within the time period when the video was captured can be estimated. This work explores the effects of frame rate harmonics that emerge when a rolling shutter based approach is used for ENF estimation from videos captured using CMOS cameras. These harmonics are a problem, especially for videos whose frame rate is a divisor of the nominal ENF because the frame rate harmonics and the ENF harmonics overlap. It is discovered that a key reason for the presence of the harmonics is the inverse square law of light that results in some repeating patterns of luminance variation across frames. This paper presents an analysis of the effect of the inverse square law of light on ENF estimation. A technique for refined ENF-related luminance signal estimation is proposed that attenuates these frame rate harmonics. This enables more accurate ENF estimates. The work also proposes an approach to estimate ENF-related luminance waveform cycles within each video frame, and a method to compute the confidence score for the estimated cycles. It provides insight into the reliability of the extracted ENF signal from a video, in the sense of its usefulness for ENF forensics, and consequently for ENF detection, which is an important precursor to ENF-based video forensics.
- 4 November 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Information Forensics and Security
- Vol. 18 (15566013) , 248-260
The Effect of Inverse Square Law of Light on ENF in Videos Exposed by Rolling Shutter
Due to a constant imbalance between demand and supply of power, ENF (Electric Network Frequency) fluctuates around a nominal value of 50 or 60 Hz. These variations in ENF cause the luminance intensity of a mains-powered light source, having no AC/DC converter inside, also to fluctuate. As a result, a video of a scene illuminated by a mains-powered light source can be used to estimate these fluctuations. As a consequence, the ENF signal within the time period when the video was captured can be estimated. This work explores the effects of frame rate harmonics that emerge when a rolling shutter based approach is used for ENF estimation from videos captured using CMOS cameras. These harmonics are a problem, especially for videos whose frame rate is a divisor of the nominal ENF because the frame rate harmonics and the ENF harmonics overlap. It is discovered that a key reason for the presence of the harmonics is the inverse square law of light that results in some repeating patterns of luminance variation across frames. This paper presents an analysis of the effect of the inverse square law of light on ENF estimation. A technique for refined ENF-related luminance signal estimation is proposed that attenuates these frame rate harmonics. This enables more accurate ENF estimates. The work also proposes an approach to estimate ENF-related luminance waveform cycles within each video frame, and a method to compute the confidence score for the estimated cycles. It provides insight into the reliability of the extracted ENF signal from a video, in the sense of its usefulness for ENF forensics, and consequently for ENF detection, which is an important precursor to ENF-based video forensics.
- 9 September 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Signal Processing Letters
- Vol. 29 (10709908) , 1963-1967
ENF Based Robust Media Time-Stamping
Electric Network Frequency (ENF) continuously fluctuates around a nominal value (50/60 Hz) due to a persistent imbalance between supplied and demanded power. In certain circumstances, ENF gets intrinsically embedded into audio and video recordings and can be extracted from these recordings. Consequently, ENF can be used in a number of media forensic applications, such as verifying the time of recording of the media. In this work, a robust media time-stamping approach is proposed for media whose ENF content is relatively contaminated. It essentially entails two procedures: first, detecting all useful, i.e., considerably accurate, samples of an estimated ENF signal, and then applying an adapted normalized cross-correlation process that is designed for exploiting just the selected ENF portions based on a binary mask of the identified accurate samples. Experimental results show that the proposed approach provides significantly increased performance.
- 9 September 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Signal Processing Letters
- Vol. 29 (10709908) , 1963-1967
ENF Based Robust Media Time-Stamping
Electric Network Frequency (ENF) continuously fluctuates around a nominal value (50/60 Hz) due to a persistent imbalance between supplied and demanded power. In certain circumstances, ENF gets intrinsically embedded into audio and video recordings and can be extracted from these recordings. Consequently, ENF can be used in a number of media forensic applications, such as verifying the time of recording of the media. In this work, a robust media time-stamping approach is proposed for media whose ENF content is relatively contaminated. It essentially entails two procedures: first, detecting all useful, i.e., considerably accurate, samples of an estimated ENF signal, and then applying an adapted normalized cross-correlation process that is designed for exploiting just the selected ENF portions based on a binary mask of the identified accurate samples. Experimental results show that the proposed approach provides significantly increased performance.
- 4 July 2022
- Published by The Academy of Transdisciplinary Learning and Advanced Studies in Transdisciplinary Journal of Engineering & Science
- Vol. 13
Fault Classification of Wind Turbine Gearbox Bearings Based on Convolutional Neural Networks
Gearbox bearings are critical elements of wind power generation systems. Their stable operation supports the power generation, thus reducing the downtime and improving the economic efficiency of wind farms. With the wide availability of sensors, data-driven methods have started to be utilized instead of physical-based methods for condition monitoring of wind energy infrastructures. Deep learning provides significant advantages to achieving this end due to its ability to extract and select representative features without expert knowledge. The present study proposed an intelligent method based on one-dimensional convolutional neural networks (1D-CNN) to extract useful features from the vibration signals and classify different bearing faults. The performance of the proposed 1D-CNN model was evaluated employing the Case Western Reserve University dataset. As a result, the proposed method achieved an average prediction accuracy of 99.56%. The findings confirmed that the method has good stability and potentially be used to reduce operation and maintenance costs.
- 4 July 2022
- Published by The Academy of Transdisciplinary Learning and Advanced Studies in Transdisciplinary Journal of Engineering & Science
- Vol. 13
Fault Classification of Wind Turbine Gearbox Bearings Based on Convolutional Neural Networks
Gearbox bearings are critical elements of wind power generation systems. Their stable operation supports the power generation, thus reducing the downtime and improving the economic efficiency of wind farms. With the wide availability of sensors, data-driven methods have started to be utilized instead of physical-based methods for condition monitoring of wind energy infrastructures. Deep learning provides significant advantages to achieving this end due to its ability to extract and select representative features without expert knowledge. The present study proposed an intelligent method based on one-dimensional convolutional neural networks (1D-CNN) to extract useful features from the vibration signals and classify different bearing faults. The performance of the proposed 1D-CNN model was evaluated employing the Case Western Reserve University dataset. As a result, the proposed method achieved an average prediction accuracy of 99.56%. The findings confirmed that the method has good stability and potentially be used to reduce operation and maintenance costs.
- 29 December 2020
- Published by Muhendislik Bilimleri ve Tasarim Dergisi in Mühendislik Bilimleri ve Tasarım Dergisi
- Vol. 8 (5) , 306-316
DERİN ÖĞRENME TABANLI VE PID KONTROL TABANLI SÜRÜCÜSÜZ ARAÇ SİSTEMLERİ
İnsan müdahalesi olmadan kendi kendine hareket edebilen araçlar sürücüsüz araç olarak adlandırılmaktadır. Sürücüsüz araçlar son yirmi yılda; askeri, lojistik ve endüstriyel üretimdeki potansiyel uygulamaları ile hem akademiden, hem de endüstriden büyük ilgi görmeye başlamıştır. Sürücüsüz araçların kullanılması günümüz trafiğinin çevresel etkilerini azaltmak ve trafik kazalarını önlemek gibi birçok konuda toplumsal fayda sağlamaktadır. Sürücüsüz araçlarda navigasyon için GPS, çarpışmaları önlemek için sensör ve nesneleri tespit etmek için kamera gibi çeşitli teknolojiler kullanılmaktadır. Bu teknolojilerin hepsi ya da birkaçı kullanılarak Derin Öğrenme tabanlı ve PID kontrol ile otonom sürüş yapılabilmektedir. Bu çalışmada Derin Öğrenme Tabanlı model eğitimi ve otonom sürüş testleri sürüş simülatöründe gerçekleştirilmiştir. Sürüş simülatöründen aracın direksiyon açısı, hız bilgisi ve ön camına monte edilen üç kameradan (sağ, sol ve orta) görüntü bilgisi alınmıştır. Aracın otonom hareketi Derin Öğrenme tabanlı model eğitimi gerçekleştirilerek ve PID kontrol ile sağlanmıştır. Bu çalışmada Derin Öğrenme ile eğitilen modelin sürüş performansı ile PID kontrol ile gerçekleştirilen sürüş performansı sürüş simülatöründe bir tam turda karşılaştırılmıştır. Aracın sürüş parkurundaki bir tam turda gerçek zamanlı olarak özerk hareketi kaydedilmiş ve başarım değerlendirmesi gerçekleştirilmiştir. Sürüş simülatöründe gerçekleştirilen testler sonucunda PID kontrol tabanlı sürüşte de başarılı sonuçlar elde edilmiş olmasına rağmen, Derin Öğrenme tabanlı modelin performansının daha iyi olduğu belirlenmiştir.
- 29 December 2020
- Published by Muhendislik Bilimleri ve Tasarim Dergisi in Mühendislik Bilimleri ve Tasarım Dergisi
- Vol. 8 (5) , 306-316
DERİN ÖĞRENME TABANLI VE PID KONTROL TABANLI SÜRÜCÜSÜZ ARAÇ SİSTEMLERİ
İnsan müdahalesi olmadan kendi kendine hareket edebilen araçlar sürücüsüz araç olarak adlandırılmaktadır. Sürücüsüz araçlar son yirmi yılda; askeri, lojistik ve endüstriyel üretimdeki potansiyel uygulamaları ile hem akademiden, hem de endüstriden büyük ilgi görmeye başlamıştır. Sürücüsüz araçların kullanılması günümüz trafiğinin çevresel etkilerini azaltmak ve trafik kazalarını önlemek gibi birçok konuda toplumsal fayda sağlamaktadır. Sürücüsüz araçlarda navigasyon için GPS, çarpışmaları önlemek için sensör ve nesneleri tespit etmek için kamera gibi çeşitli teknolojiler kullanılmaktadır. Bu teknolojilerin hepsi ya da birkaçı kullanılarak Derin Öğrenme tabanlı ve PID kontrol ile otonom sürüş yapılabilmektedir. Bu çalışmada Derin Öğrenme Tabanlı model eğitimi ve otonom sürüş testleri sürüş simülatöründe gerçekleştirilmiştir. Sürüş simülatöründen aracın direksiyon açısı, hız bilgisi ve ön camına monte edilen üç kameradan (sağ, sol ve orta) görüntü bilgisi alınmıştır. Aracın otonom hareketi Derin Öğrenme tabanlı model eğitimi gerçekleştirilerek ve PID kontrol ile sağlanmıştır. Bu çalışmada Derin Öğrenme ile eğitilen modelin sürüş performansı ile PID kontrol ile gerçekleştirilen sürüş performansı sürüş simülatöründe bir tam turda karşılaştırılmıştır. Aracın sürüş parkurundaki bir tam turda gerçek zamanlı olarak özerk hareketi kaydedilmiş ve başarım değerlendirmesi gerçekleştirilmiştir. Sürüş simülatöründe gerçekleştirilen testler sonucunda PID kontrol tabanlı sürüşte de başarılı sonuçlar elde edilmiş olmasına rağmen, Derin Öğrenme tabanlı modelin performansının daha iyi olduğu belirlenmiştir.
- Open Access
Fusion of Camera Model and Source Device Specific Forensic Methods for Improved Tamper Detection
PRNU based camera recognition method is widely studied in the image forensic literature. In recent years, CNN based camera model recognition methods have been developed. These two methods also provide solutions to tamper localization problem. In this paper, we propose their combination via a Neural Network to achieve better small-scale tamper detection performance. According to the results, the fusion method performs better than underlying methods even under high JPEG compression. For forgeries as small as 100$\times$100 pixel size, the proposed method outperforms the state-of-the-art, which validates the usefulness of fusion for localization of small-size image forgeries. We believe the proposed approach is feasible for any tamper-detection pipeline using the PRNU based methodology.
Open AccessFusion of Camera Model and Source Device Specific Forensic Methods for Improved Tamper Detection
PRNU based camera recognition method is widely studied in the image forensic literature. In recent years, CNN based camera model recognition methods have been developed. These two methods also provide solutions to tamper localization problem. In this paper, we propose their combination via a Neural Network to achieve better small-scale tamper detection performance. According to the results, the fusion method performs better than underlying methods even under high JPEG compression. For forgeries as small as 100$\times$100 pixel size, the proposed method outperforms the state-of-the-art, which validates the usefulness of fusion for localization of small-size image forgeries. We believe the proposed approach is feasible for any tamper-detection pipeline using the PRNU based methodology.
- 1 September 2019
- Published by Elsevier in Digital Investigation
- Vol. 30, 43-51
PRNU based source camera attribution for image sets anonymized with patch-match algorithm
- 1 September 2019
- Published by Elsevier in Digital Investigation
- Vol. 30, 43-51
PRNU based source camera attribution for image sets anonymized with patch-match algorithm
- 1 June 2019
- Published by Elsevier in Digital Investigation
- Vol. 29, 91-100
PRNU-based source device attribution for YouTube videos
- 1 June 2019
- Published by Elsevier in Digital Investigation
- Vol. 29, 91-100
PRNU-based source device attribution for YouTube videos