Semester projects I have supervised within IVRL. More details about the projects can be found here, this list only contains the summaries.
* Last update 03/2018 *
“Next Big Hit?” (1) (SPRING 2018)
Objective: Study music features and user data to build a novel recommendation system that does not optimize maximum likelihood of a user listening to a given song on their own, but a different novel metric.
“Next Big Hit?” (2) (SPRING 2018)
Objective: Predict whether a song will be popular for the general public by studying music features and user data to solve a sub-problem, namely to first predict whether a given category of users will find a song popular.
“Exploring Beautiful Scenes by Location” (1) (FALL 2017)
Objective: Develop a website for exploring photos geographically on a map. The entire infrastructure along with user-action tracking, user profiles and search functions.
“Exploring Beautiful Scenes by Location” (2) (SPRING 2018)
Objective: Improve the design of the front end of the photo exploration website. Develop three types of gamification on top of the tool to encourage user engagement. Simulate the evolution of the gamification.
“Pushing the Limits of Secret Messages in Images” (FALL 2017)
Objective: Implement a novel algorithm that makes the steganalysis a lot more complex by adding a maximum of randomness to the process, while maintaining a correct decoding. The steganography is done spatially, with primary objective “an impossible adversarial decoding” and secondary objective “a hard to detect image modification”.