Okan Köpüklü

About Me

I am a researcher in Microsoft’s Applied Science Group (ASG). Previously, I worked on my Ph.D. project between 2017-2021 at Institute of Human-Machine Communication, Technical University of Munich (TUM), supervised by Prof. Gerhard Rigoll. Before that, I received my Master's degree in Electrical Engineering from Technical University of Munich and my Bachelor's degree from Boğaziçi University in 2017 and 2014, respectively.

My research interests lie in computer vision, audio and audiovisual signal processing, pattern recognition, deep learning, and their applications.

Education

Nov. 2017 - Nov. 2021, Department of Electrical and Computer Engineering, Technical University of Munich, Germany,

Ph.D.

Oct. 2014 - Feb. 2017, Department of Electrical and Computer Engineering, Technical University of Munich, Germany,

Master of Science.

Sep. 2011 - Jun. 2014, Department of Electrical and Electronics Engineering, Boğaziçi University, Istanbul, Turkey,

Bachelor of Science.

Work Experience

Microsoft
Researcher
Nov. 2021 - Present, Munich, Germany
integrating state-of-the-art deep learning algorithms into modern devices.

Microsoft
Research Internship
June 2021 - Aug. 2021, Munich, Germany
Real-time audio-visual active speaker detection.
Resource efficient architecture design for voice activity detection.

RYSTA GmbH
Software Developer
Apr. 2017 - Nov. 2017, Munich, Germany
Developement of lightweight machine learning algorithms for IoT devices.

BMW Group AG
Master Thesis Student
Jun. 2016 - Dec. 2016, Munich, Germany
Modeling of in-car communication traffic.

Intel Labs Munich
Working Student
Jan. 2015 - Jun. 2016, Munich, Germany
Time series modeling for energy grid power consumption.

Publications

• Dissected 3D CNNs: Temporal Skip Connections for Efficient Online Video Processing, in Computer Vision and Image Understanding 2022
  Accepted to Journal of Computer Vision and Image Understanding!
  O. Köpüklü, S. Hörmann, F. Herzog, H. Cevikalp, G. Rigoll   [code]
  


• How to Design a Three-Stage Architecture for Audio-Visual Active Speaker Detection in the Wild, in ICCV 2021 O. Köpüklü, M. Taseska, G. Rigoll   [code]
  


• Deep Compact Polyhedral Conic Classifier for Open and Closed Set Recognition, in Pattern Recognition 2021 H. Cevikalp, B. Uzun, O. Köpüklü, G. Ozturk   [code]
  


• U-Net based Zero-hour Defect Inspection of Electronic Components and Semiconductors, in VISIGRAPP 2021
  F. Kälber, O. Köpüklü, N.H. Lehment, G. Rigoll  


• Driver Anomaly Detection: A Dataset and Contrastive Learning Approach, in WACV 2021
  O. Köpüklü, J. Zheng, H. Xu, G. Rigoll   [dataset, code]
  


• DriverMHG: A Multi-Modal Dataset for Dynamic Recognition of Driver Micro Hand Gestures and a Real-Time Recognition Framework, in FG 2020
  O. Köpüklü, T. Ledwon, Y. Rong, N. Kose, G. Rigoll   [dataset, code]
  


• Online Dynamic Hand Gesture Recognition Including Efficiency Analysis, in TBIOM 2020
  O. Köpüklü, A. Gunduz, N. Kose, G. Rigoll   [code]
  


• Real-time hand gesture detection and classification using convolutional neural networks, in FG 2019
  O. Köpüklü, A. Gunduz, N. Kose, G. Rigoll   [code]
  Best Student Paper Award at FG 2019!
  


• Resource efficient 3d convolutional neural networks, in ICCV Neural Architects Workshop 2019
  O. Köpüklü, N. Kose, A. Gunduz, G. Rigoll   [code]
  


• Talking with your hands: Scaling hand gestures and recognition with cnns, in ICCV Hands Workshop 2019
  O. Köpüklü, Y. Rong, G. Rigoll   [dataset], [code]
  


• You Only Watch Once: A Unified CNN Architecture for Real-Time Spatiotemporal Action Localization, in arxiv 2019
  O. Köpüklü, X. Wei, G. Rigoll   [code]
  


• Convolutional neural networks with layer reuse, in ICIP 2019
  O. Köpüklü, M. Babaee, S. Hörmann, G. Rigoll   [code]
  


• Deep Attention Based Semi-Supervised 2D-Pose Estimation for Surgical Instruments, in arxiv 2020
  M. Kayhan, O. Köpüklü, M. H. Sarhan, M. Yigitsoy, A. Eslami, G. Rigoll   [code]
  


• Unsupervised Monocular Depth Prediction for Indoor Continuous Video Streams, in arxiv 2020
  Y. Feng, S. Wu, O. Köpüklü, X., F. Tombari  


• Real-Time Driver State Monitoring Using a CNN Based Spatio-Temporal Approach, in ITSC 2019
  N. Kose, O. Köpüklü, A. Unnervik, G. Rigoll  


• Gait Energy Image Restoration Using Generative Adversarial Networks, in ICIP 2019
  M. Babaee, Y. Zhu, O. Köpüklü, S. Hörmann, G. Rigoll  


• Outlier-Robust Neural Aggregation Network for Video Face Identification, in ICIP 2019
  S. Hörmann, M. Knoche, M. Babaee, O. Köpüklü, G. Rigoll  


• Comparative Analysis of CNN-based Spatiotemporal Reasoning in Videos, in arxiv 2019
  O. Köpüklü, F. Herzog, G. Rigoll   [code]
  


• Motion fused frames: Data level fusion strategy for hand gesture recognition, in CVPR AMFG Workshop 2018
  O. Köpüklü, N. Kose, G. Rigoll   [code]
  


• Analysis on Temporal Dimension of Inputs for 3D Convolutional Neural Networks, in IPAS 2018
  O. Köpüklü, G. Rigoll  

Professional Service

• Teaching assistant of Pattern Recognition course (Master's level), 2019, 2020, 2021
• Reviewer of IEEE TNNLS 2021
• Reviewer of IEEE TCSS 2021
• Reviewer of IEEE JBHI 2021
• Reviewer of WACV 2021
• Reviewer of BMVC 2020, 2021
• Reviewer of IEEE TCAS-I 2020
• Reviewer of ITSC 2019
• Reviewer of EURASIP JIVP 2018

Honors and Awards

• 2^nd place at ActivityNet Challenge, AVA Active Speaker Detection Task.   2021
• Best student paper award in FG’19.   2019
• Graduation with exception from Master's degree (TUM).   2017
• Graduation with high honor degree from Bachelor's degree (BOUN).   2014