About Me

Welcome to my personal webpage. I currently work at Zoox as a Research Engineer in the Vision team. I got my PhD degree at Vision Research Lab in UCSB, with Prof. B.S. Manjunath as my advisor. My research area is Computer Vision and Machine Learning.

Professional Experience

  • 1/2020 - Present: Zoox, Computer Vision Engineer

  • 6/2015 - 12/2019: Vision Research Lab, Graduate Student Researcher

  • Conference Reviewer: CVPR - 2018 , ACCV - 2018, Transactions on Multimedia

  • 6/2018 - 9/2018: IBM Research, Summer Research Intern

  • 6/2017 - 9/2017: IBM Research, Summer Research Intern

  • 1/2017 - 3/2017: Army Research Lab, Research Intern

  • 7/2016 - 8/2016: Army Research Lab, Research Intern

  • 6/2015- 8/2015: Vision Research Lab, Student Mentor. Projects Page

  • 1/2015 - 8/2015: ECE Department, Teaching Assistant

Publications

  • O. Ulutan*, A.S.M. Iftekhar*, B.S. Manjunath, “VSGNet: Spatial Attention Network for Detecting Human-Object Interactions using Graph Convolutions”, In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2020, Seattle, WA, 2020. Arxiv

  • O. Ulutan, S. Rallapalli, C. Torres, M. Srivatsa, and B. Manjunath, "Actor Conditioned Attention Maps for Video Action Detection", In Proceedings of the IEEE Winter Conference on Applications of Computer Vision (WACV) 2020, Colorado, USA, 2020. Arxiv

  • O. Ulutan, B.S. Riggan, N.M. Nasrabadi, B.S. Manjunath, "An Order Preserving Bilinear Model for Person Detection in Multi-Modal Data", Winter Conference on Applications of Computer Vision, (WACV 2018), Lake Tahoe, USA, March 2018. Arxiv

  • S. Kumar, C. Torres, O. Ulutan, A. Ayasse, D. Roberts, B.S. Manjunath. “Deep Remote Sensing Methods for Methane Detection in Overhead Hyperspectral Imagery”, In Proceedings of the IEEE Winter Conference on Applications of Computer Vision (WACV) 2020, Colorado, USA, 2020. Link

  • Liu, Xiaochen; Ulutan, Oytun; Chan, Kevin; Manjunath, B.S.; Govindan, Ramesh; "Caesar: Cross-camera Complex Activity Detection", ACM SenSYS 2019.

  • Celso de Melo*, Brandon Rothrock, Prudhvi Gurram, Oytun Ulutan, B.S. Manjunath, "Vision-Based Gesture Recognition in Human-Robot Teams Using Synthetic Data", 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems, Las Vegas, USA, October 2020.

  • ShreeRanjani SrirangamSridharan, Oytun Ulutan, Shehzad Noor Taus Priyo, Swati Rallapalli, Mudhakar Srivatsa, "Object Localization and Size Estimation from RGB-D Images", Arxiv, Pre-print 2018. Arxiv

  • Ustunkaya, Tuna, Benoit Desjardins, Bolun Liu, Jaeseok Park, Oytun Ulutan, Nissi Saju, Francis E. Marchlinski, and Saman Nazarian. "Association of Regional Myocardial Conduction Velocity With the Distribution of Endocardial Hypoattenuation on Contrast-Enhanced CT in Patients With Post-Infarct Ventricular Tachycardia Substrate." Circulation 138, no. Suppl_1 (2018): A10738-A10738.

  • V. Acikel, O. Ulutan, A. C. Ozen, B. Akin, Y. Eryaman, E. Atalar, "A Novel MRI Based Electrical Properties Measurement Technique". International Society for Magnetic Resonance in Medicine (ISMRM 2013).), Salt Like City, USA, April 2013.

Workshops/Challenges

  • Oytun Ulutan, B. S. Manjunath. “UCSB, ActivityNet-AVA Actions Challenge Solution Actor Conditioned Attention Maps for Video Action Detection”. UCSB Submission to ActivityNet AVA - CVPRW 2019

  • Oytun Ulutan, Swati Rallapalli, Mudhakar Srivatsa, B.S. Manjunath. “Joint Classification and Detection Using LSTMs” UCSB & IBM Research Submission to ActivityNet Challenge - CVPRW 2017, Notebook Paper Link,

Projects

Pedestrian Attributes and Gestures

Classification of pedestrian attributes and gestures from cameras on self driving vehicles.

Project at Zoox

lab_actions_test.mp4

Actor Conditioned Attention Maps for Video Action Detection

Authors: Oytun Ulutan, Swati Rallapalli, Mudhakar Srivatsa, B.S. Manjunath

In collaboration with IBM Research.

Real-time action detection demo is available on Github.

Training code available on Github.

Model running on various surveillance videos is available at video link.

Paper Link: Arxiv

VSGNet: Spatial Attention Network for Detecting Human Object Interactions Using Graph Convolutions

Authors: Oytun Ulutan*, ASM Iftekhar*, B.S. Manjunath

Achieved state of the art results on V-COCO and HICO-det Datasets!

Code is available on Github.

Paper Link: Arxiv

visdrone_uav_1_output.mp4

Interaction Detection on Drone Footages

In collaboration with Army Research Lab and USC.

Code for Object Detector on drone views is available on Github.

More details coming soon.

Caesar: Cross-camera Complex Activity Detection

Authors: Liu, Xiaochen; Ulutan, Oytun; Chan, Kevin; Manjunath, B.S.; Govindan, Ramesh

Published in ACM SenSYS 2019.

In collaboration with USC and Army Research Lab.

More details coming soon.


Joint Classification and Detection Using LSTMs

UCSB & IBM Research Submission to ActivityNet Challenge 2017

Authors: Oytun Ulutan, Swati Rallapalli, Mudhakar Srivatsa, B.S. Manjunath

In collaboration with IBM Research.

Notebook Paper Link, also included in Challenge Summary

An Order Preserving Bilinear Model for Person Detection in Multi-Modal Data

Authors: Oytun Ulutan, Benjamin S. Riggan, Nasser M. Nasrabadi, B.S. Manjunath

Published in WACV2018.

In collaboration with Army Research Lab.

Paper Link: Arxiv

Code: Github

Poster - Spotlight Presentation - Spotlight Video

Undergraduate/Masters Projects

  • Smart Auditive Device: A device that is designed to notify hearing-impaired people of surrounding voices and sounds, using speech processing and machine learning Methods. (VHDL, C)

Received a funding offer for this project from Turkish Ministry of Industry.

  • Pedestrian Detection using a Camera: This system uses multiple cascade classifiers on multiple features (HOG, HAAR) to detect the pedestrian presence and orientation from the live feed. It has been implemented on both C++ and Android. OpenCV is widely used in both implementations.

Supervised by Prof. B.S. Manjunath.

  • Isolated Digit Recognizer: Speech recognition project that recognizes isolated digits from the recorded speech. This system uses Mel Frequency Cepstral Coefficients as features and detects the digit in the recorded speech.

Supervised by Prof. Lawrence Rabiner.

  • Smart Parking Lot Routing System: Wireless Networking Technologies course project. With three access points in an outdoor parking lot, user’s location is found using localization algorithms. After the user is located, he is guided to the empty parking space closest to his location. (Android)

Supervised by Prof. Ezhan Karasan.

  • Economic Data analysis: A course related project that consists of examining stock market and designing a method to predict future events. (Matlab)

Supervised by Prof. Serdar Kozat.

  • Speech Processing Project: A course related project that consists of obtaining a speech, classifying it and responding according to the speech. (VHDL, Analog Microphone Design)

Supervised by Prof. Tolga Mete Duman.

  • Digital Electronics, Phase detection system: This system determines the phase of an input signal with respect to another signal using a FPGA and a microprocessor. (VHDL, C)

Supervised by Prof. Ziya Ider, Presented to ABET as an example project.

  • Digital Design Project, Digital Thermometer: This system determines the temperature of the room using a sensor driven by a FPGA. (VHDL)

Supervised by Prof. Ergin Atalar

  • Programming Project, Panoramic Map: Designed a program which displays panoramic photos of Bilkent University campus using Java. (Java)

Supervised by Prof. Hakan Ferhatosmanoglu