Direct Interaction and User-defined Saliency for 360° Video Navigation in Virtual Reality
Currently conducting research at DGP, working with professor Tovi Grossman and postdoctoral fellow Ludwig Sidenmark, in what 360° video navigation in VR could be when the user is given full control to define what is of importance to them and directly interact with these "objects" of interest during playback. This is in contrast to the research done in this field that mostly focuses on guiding the user such that their gaze aligns with what the author depicts as important.
As part of the Real-Time Production R&D Engineering team I worked on really awesome stuff such as using ML for 3d environment reconstruction and developing new features for the proprietary 3d content creation software primarily focused on real-time rendering. While working with C++, I played a role in developing the rendering API, and for the ML project in Python, I utilized PyTorch as the main framework.
I led the testing and validation of the Cybertruck's powertrain firmware, specifically focusing on differential lockers in off-road modes. I designed, implemented, and executed Hardware-in-the-Loop (HIL) tests to ensure comprehensive coverage of all specified requirements. Additionally, I conducted in-depth analysis and evaluation of ECU data, ensuring optimal performance and reliability.
Working at Nuwa, home to the world’s first pen that digitizes human writing on any paper surface, I created an autonomous testing unit simulating human’s note-taking, content and motion, to validate the engineering design. Using complex path planning concepts, and a text-to-handwriting generation ML model, I was able to recreate a human-pen interaction in a controlled environment in order to ensure thorough testing and achieve the highest quality assurance possible.
While working at Finn Land Research Lab I helped cognitive neuroscience researchers by creating python scripts to carry out their requested data analysis. My program would interact with a large number of excel files making use of the information stored in the cells and analyze it, storing the findings for the researchers.
My work at Centivizer consisted of designing consumer products using SolidWorks and gathering data from finite element analysis, FEA, using ANSYS. In addition, I was in-charge of market research to find the optimal mold-injection manufacturer for our products and the best heart rate tracker to implement into a product.