Project: #18 Dynamic 3D Reconstruction of Vehicles for Safer Intersections Progress Report - Reporting Period Ending: March 31, 2018 Principal Investigator: Srinivasa Narasimhan Status: Completed Start Date: Jan. 1, 2017 End Date: Aug. 31, 2018 Research Type: Advanced Grant Type: Research Grant Program: MAP-21 TSET National (2013 - 2018) Grant Cycle: 2017 TSET UTC Progress Report (Last Updated: July 18, 2018, 7:14 a.m.) % Project Completed to Date: 65 % Grant Award Expended: 10 % Match Expended & Document: 100 USDOT Requirements Accomplishments Significant progress in the hardware, software, and algorithm development for 3D stereoscopic headlights. Algorithms developed for 1) pedestrian detection during the night time, 2) significantly improved detection, localization and reconstruction of moving vehicles, even in the presence of strong occlusions, and 3) stereo matching objects of difficult materials with cross-spectral imaging. Many demos to visitors of NREC, UTC, industry, and academia. Presented work at Ford's Advanced Lighting Innovation Expo. Pitched the project as a potential startup at the University Technology Showcase during the 3 Rivers Venture Fair (won first place). Impacts Interactions during presentations and demonstrations to academics, companies, and state transportation agencies have been very positive in terms of informing people of the technological possibilities of solving first hand problems. Other Provisional patents in preparation for submission. Websites: http://www.cs.cmu.edu/smartheadlight/ http://www.cs.cmu.edu/~ILIM/projects/IM/CarFusion/ http://www.cs.cmu.edu/~ILIM/projects/AA/RGBNIRStereo/ Outcomes New Partners NA Issues NA