Neural Radiance Fields research and application

Research into expanding neural radiance fields and then applying it to the real world.

Thesis project description

We are looking for a project

A bachelor project centered around machine learning

Specifically something within the field of deep learing, and we welcome you to propose any projects that could benefit from our set of skills.

Currently, we have been looking at Neural Radiance Fields (NeRF) and its applications. It is a new technique that enables novel view synthesis from just a few images. This allows you to "fly" around a scene and see it from any direction. Specifically BaRF, Mip-NeRF 360, and InstantNGP have caught our eyes, and we would love to apply them to a real-world problem. This is where you could come in to help ground our work in reality.

When do we start?

We are targetting summer 2023

Our plan is to start work on the project at the end of August 2023 and finish by the end of December 2023.

The time schedule can of course be moved slightly by a few weeks or a month, but we would like to keep it as close to this as possible.

What we ask of you

A collaboration works both ways

For a collaboration to work well, we need to ensure that we are a good fit for each other. We need to ensure we can communicate and work well together.

• The project will likely require a dataset. In the case of NeRF, the dataset requirement is not as strict, since we can quickly collect our own data. 
  In the case of other projects, we will not accept collaborations that do not already have a dataset ready. This is to avoid situations where the data is unavailable for half of the project because of unforeseen dependencies.
• We want to share our methods and results with people. Therefore any NDA we sign may not be time unlimited when it comes to the methods and results. The data itself and proprietary tools/techniques can of course still be covered by an NDA.
• Atleast one representative from the company needs to be available for regular meetings. We would therefore like for you to allocate at least 1 weekly 1 hour meeting throughout the project. This can of course be discussed, but we encourage it to ensure we work towards the same goal.

Project website

We made a website with additional information

https://www.proton.run/bachelor-project

Managerial value of the research

Neural Fields, in general, is a very new and promising research field that has applications within many problem domains.Neural Radiance Fields has state of the art applications within computer vision. It enables good and cheap solutions to previously hard problems.Exposure of your organization to recent machine learning techniques to gain a competitive advantage. We can learn from each other in our meetings.Lead instead of follow. Gain a competitive advantage by being first to market with a solution based upon the newest research and capabilities.We are an independent team that requires little to no guidance to progress. The collaboration will therefore cost you very little.

Why collaborate with us

Check out our project website for more information about the team and its individual members:https://www.proton.run/bachelor-project

Researcher bio

My interests are broad and many. The general gist is that if a topic is technically deep, useful, and accessible - I will find it interesting and fun. I have partaken in various case competitions and hackathons, so I know how to work fast in a team, under pressure, and to achieve a goal. The Excelerate text field is too restrictive and cannot contain my interests and hobbies. Check out our project website that we made for this occasion. It includes much more information about me. https://www.proton.run/bachelor-project

Passion and motivation

Currently my interests are mostly in machine learning: I want to dive deeper into deep learning, computer vision, time series, and stochastic processes.