3D Computer Vision Researcher

Team Description:

Kitware's computer vision team is a leader in the creation of cutting-edge algorithms and software for automated image and video analysis. Our solutions embrace deep learning and add measurable value to government agencies, commercial organizations, and academic institutions worldwide. We understand the difficulties in extracting, interpreting, and utilizing information across images, video, metadata, and text, and we recognize the need for robust, affordable solutions. We seek to advance the fields of computer vision and deep learning through research and development and through collaborative projects that build on our open source software platforms, such as VIAME and Telesculptor.

About the Projects:

Our 3D computer vision team develops algorithms and open source applications for photogrammetry: reconstructing 3D models of objects from images and video. Current 3D projects have an emphasis on aerial and satellite imagery sources, but capabilities are often adapted to other problem domains. These projects have created and continue to support the development of TeleSculptor, an open source application for aerial 3D reconstruction built on KWIVER, and Danesfield, an open source Python framework and web application for urban 3D modeling from satellite imagery. Our 3D vision research projects support the continued development of these tools, development and publication of new cutting edge 3D reconstruction and camera calibration algorithms, and opportunities to leverage 3D vision to assist with broader computer vision topics. The team is exploring how 3D vision algorithms can aid in object detection and tracking, change detection, super resolution, video compression, super resolution, and various other problems. We frequently explore opportunities to adapt our work to new customers with new challenges and new data domains such as underwater imagery or synthetic aperture radar.

Kitware’s employees have unique opportunities to interact and collaborate directly with customers, visit interesting customer sites, and participate in live field tests and demonstrations. Much of Kitware’s work involves applying state-of-the-art artificial intelligence approaches to dynamic, real-world problems. In this case, you will have the opportunity to contribute your skills to projects focused on national security, making a difference on a daily basis to protect our country. Research and Development Engineers at Kitware also enjoy benefits commonly associated with a position in academia, such as support and encouragement for the publication of novel work.

We partner with premier government R&D agencies such as DARPA, IARPA, AFRL, NVESD, NOAA, ONR, other branches of the US military, and multiple members of the Intelligence Community on a range of efforts including prime contracts, SBIRs, and STTRs. In addition, we provide commercial services to companies ranging from startups to Fortune 500 companies. Kitware employs an open source business model to foster extended, collaborative communities, and to provide flexible, high-quality technical solutions. If you’ve used CMake, ITK, or VTK, you know our work and the impact it has on the communities we help build.

In This Position You Will:

• Conduct research and develop solutions for problems related to camera calibration, registration, and 3D reconstruction from images
• Contribute to the development of open source software such as TeleSculptor, KWIVER, or Danesfield
• Collaborate with other researchers to brainstorm ideas and peer review work
• Prepare and present research findings to customers and collaborators
• Enjoy support and encouragement for participation in national and international conferences (such as CVPR, ICCV, ECCV, WACV, and 3DV)
• Be encouraged to seek funding to grow and develop your own research areas, if you desire

Required Qualifications

• PhD in Computer Science or related field
• Experience using multiview geometry to develop algorithms for building 3D point clouds or meshes from imagery or video
• Experience with AL/ML methods for 3D reconstruction or camera calibration
• Experience with good software development practices and proficiency in C++ and/or Python
• Strong publication record in top-tier research publications and conferences such as CVPR, ICCV, and ECCV
• Highly innovative and demonstrated track record for solving difficult technical challenges using imagery
• Experience collaborating successfully with others and thriving in a fast-paced and dynamic work environment
• Candidates should include a detailed list of publications as part of their resume/CV
• Due to contractual restrictions, only US Citizens will be considered for this position
• If not already cleared TS/SCI, willingness and ability to apply for and maintain a TS/SCI security clearance

Preferred Qualifications:

• Experience with geospatial coordinates systems and transformation is desirable, but not required