Explore my research-driven portfolio
View all projects →Serhan Narlı
Research Scientist and PhD Candidate at Charité – Universitätsmedizin Berlin
I develop computer-vision and machine-learning systems for biomechanical analysis. My current work on OrthoPose focuses on patient-specific 3D pelvis, hip and knee kinematics from monocular video, while my broader research spans animal gait analysis, clinical motion assessment and production-ready AI pipelines.
About
I am a Research Scientist and PhD candidate at Charité – Universitätsmedizin Berlin. My work combines computer vision, machine learning and biomechanics to derive clinically meaningful movement information from video.
In OrthoPose, I investigate patient-specific three-dimensional pelvis, hip and knee kinematics from monocular recordings. The methodology integrates image-derived motion, anthropometric information, anatomical constraints and clinical motion-capture supervision.
My broader background includes animal gait analysis, medical imaging, signal processing, edge AI and the translation of research prototypes into reliable software systems.
Research and publications
My research focuses on video-based movement analysis, interpretable biomechanical features and clinically relevant validation.
A computer-vision pipeline for analysing bovine back shape and extracting curvature-based features to distinguish sound and lame animals in comparison with human locomotion scores.
Publication link coming soonA validation study comparing deep-learning gait assessment with multiple human raters and examining the conditions under which automated scoring can match or exceed human consistency.
Publication link coming soonA CNN-based approach to COVID-19 recognition on enhanced chest radiographs, including dedicated preprocessing and architecture variants for improved diagnostic performance.
Publication link coming soonSelected projects
Research prototypes and applied AI systems connecting scientific methodology with real-world deployment.
A web-based platform that analyses walking videos, detects anatomical keypoints, derives curvature and gait features, and estimates lameness-related patterns.
A farm-ready NVIDIA Jetson pipeline combining camera acquisition, GStreamer, object detection and data-quality monitoring for real-time gait analysis.
A conversational AI platform using retrieval-augmented generation, prompt engineering and strict persona constraints to support historically grounded dialogue.
A deep-learning pipeline for detecting COVID-19 on enhanced chest X-rays using dedicated preprocessing and CNN architectures.
Writing and essays
Articles on neuroscience, artificial intelligence and the relationship between technology and society.
A long-form essay on the development of EEG from early electrical experiments to modern quantitative analysis and brain–computer interfaces.
Read on Medium →An accessible discussion of learning, perceptrons, neural networks and the role of error from both technical and philosophical perspectives.
Read on Medium →A critical examination of what modern brain–machine interfaces can realistically achieve and where public expectations exceed current evidence.
Read on Medium →An article on EEG acquisition, signal processing and deep-learning methods, informed by an Erasmus research project on emotion recognition.
Read on Medium →Experience
Research Scientist
Charité – Universitätsmedizin Berlin
Working on OrthoPose, a research project for patient-specific 3D pelvis, hip and knee kinematics from monocular video. I develop computer-vision and anatomy-constrained machine-learning methods that combine video, anthropometry and clinical motion-capture supervision.
AI/ML Engineer
MCG Motion Capture GmbH, Berlin
Developed production-ready computer-vision features, Jetson-based pilots and AI integrations for commercial motion-capture products.
Signal Processing Engineer Intern
StepUp Air Solutions, Copenhagen
Built signal-processing pipelines for wearable respiratory sensors and real-time monitoring.
Research Intern
Wroclaw University of Science and Technology
Worked on computer-vision and autonomous-systems prototypes in international research collaborations.
Education
Charité – Universitätsmedizin Berlin
Deep learning for biomechanical analysis and early lameness detection, combining video-based gait assessment with clinical validation.
İskenderun Technical University
Thesis on deep learning for medical image analysis and biomedical signal processing.
İskenderun Technical University
Development of an EEG-controlled six-axis robotic arm and embedded-systems projects.