Si Lu

Computer Vision Engineer - Autonomous Vehicles at NVIDIA, CA
Ph.D.in Department of Computer Science Maseeh College of Engineering and Computer Science Portland State University, OR

Email: lusi "at" pdx.edu
Resume (updated on 01-04-2019)

Working on Autonomous Driving at NVIDIA. PhD student in Computer Science at Portland State University. My advisor is Dr. Feng Liu. I have More than 9 years' experience on Computer Vision & Graphics topics such as 3D construction/enhancement in light field VR, novel view synthesis for camera arrays, video stabilization, image stitching as well as computational photography such as image/RGBD denoising and patch matching. Passionate about developing/implementing state-of-the-art computer vision and computer graphics algorithms. Skilled programmer with C/C++/Python/Matlab/OpenCV/OpenGL/Pytorch. Familiar with machine learning and deep learning techniques such as CNN/RNN/LSTM/ResNet/DenseNet/Faster R-CNN/Mask R-CNN.

Enducation

Portland State University	Ph.D. candidate in Computer Science	GPA 3.92	2012-present
Portland State University	Master of Computer Science	GPA 3.92	2012-2019
Tsinghua University	Master of Engineering in Fluid Mechanics	GPA 3.63	2009-2012
Tsinghua University	Bachelor of Engineering in Mechanics	GPA 3.61	2005-2009

Publications

• Si Lu, High-speed Video from Asunchronous Camera Array. Video demo (New!). Poster (New!).
  IEEE WACV 2019, Hawaii, HI, January 2019.

• Si Lu, Good Similar Patches for Image Denoising. Video demo (New!). Poster (New!).
  IEEE WACV 2019, Hawaii, HI, January 2019.

• Si Lu, No-reference Image Denoising Quality Assessment.
  Computer Vision Conference, CVC 2019, Las Vegas, NV, April 2019 (New!).

• Si Lu, Xiaofeng Ren, and Feng Liu. Depth Enhancement via Low-rank Matrix Completion.
  IEEE CVPR 2014, Columbus, OH, U.S., June 2014. Project website. Poster.

• Si Lu, Shuai Zhao, and Li Liu. Field Line Detection Based on Local-precise Extracting and Modified Hough Transform. IEEE ICVRCA 2012, Guangzhou, China, December 2012. Paper.

Experience

Computer Vision Engineer - Autonomous Vehicles

NVIDIA, Santa Clara, CA, 06/2018 - present

                                                   I am proud that my work appeared in the GTC keynote demo at GTC 2019, San Jose, CA

Autonomous Vehicles: HD Map perception

• HD map land marking fusion: lanes, poles, signs via adaptive clusters
• HD map visualization in autonomous vehicle roadrunner rendering
• Transfering to localization team to develop AV localization algorithms (new!)

Computer Vision Intern - Lytro Immerge (VR)

Lytro, Moutain View, CA, 06/2017 - 09/2017

Multi-view 3D reconstruction ( C/C++/OpenCV )
• Processing HD (2048×2048) RGB-D images
• Signigicant fine detal improvements for VR depth maps (left)
• Depth enhancement for light filed (multi-view) camera arrays (middle)
• Depth enchancement for stiched 360-degree videos in Virtual Reality (right)
• Integrated into Lytro Immerge (VR) production pipeline for depth

Research Projects at Universities
	Highspeed Videos from Camera Arrays (2016-2017) • Develope an algorithm to generate highspeed videos from camera arrays • Use novel views synthesis to do frame interpolation among cameras  • Use image-based rendering techniques, e.g., feature matching/warping • Plausible and parallax-free novel views can be robustly interpolated • Works for challenging scenes with large camera and object motions • Accepted to WACV 2019: Paper, Video demo, Poster.
	Patch Matching for Image Denoising (2017-present) Develope a clustering-based approach with unreliable pixel estimation to consistently improve patch-based denoising techniques' (like BM3D) performance via better similar patch searching for image denoising. (Accepted to WACV 2019: Paper, Video demo, Poster)
	No-reference Image Denoising Quality Assessment (2015-2017) Present a noreference image denoising quality assessment method that can be used to select for an input noisy image the right denoising algorithm with the optimal parameter setting. Paper.
	Depth Enhancement (2012-2014) Develop a depth map enhancement algorithm that performs depth map completion and denoising simultaneously for RGBD cameras like Microsoft Kinect and Xtions Pro. Paper. Project website. Poster.
	Flash Light Detection for Unmanned Aerial Vehicles (2015) Developed an algorithm that allows real-time flash light detection for Unmanned Aerial Vehicles. By designing specific color-based features, we achieved a 89.2% detection rate with a small false positive rate (0.6%).
	Robocup Team Leader (2007-2012) Project website. • Developed a robot vision system with accurate and real-time (20 FPS) objects (football, field lines and goals) detection and ball-locating/self-locating (C++/Windows Mobile SDK/Linux). • Developed a wireless vision parameter tuning interface to allow fast camera calibration before real games. By utilizing this interface, I reduced the parameter tuning time by more than 70% before each game (Matlab, C).

Teaching

CS 447/547-Computer Graphics (instructor): teach Computer Graphics topics on movies, games, animations and 3d rendering. Helped students with two course projects - a mini-Photoshop (FLTK/OpenCV) with basic image processing and an Amusement Park 3D animation rendering project (OpenGL).

CS 410/510-Computational Photography (instructor): teach research topics ranging from concepts of digital camera and photography to computer vision/graphics techniques, including high dynamic range imaging, panorama stitching, image segmentation & matting, video stabilization, virtual reality basics, deep learning in computer vision etc..