Tobias Lømo

• Lømo, Tobias; Baselizadeh, Adel; Ellefsen, Kai Olav & Tørresen, Jim (2026). Dual Process Dreamer: Fast and Slow Decision-Making with World Models. Proceedings of the International Conference on Agents and Artificial Intelligence (ICAART). ISSN 2184-3589. 2, s. 1230–1241. doi: 10.5220/0014243200004052. Fulltekst i vitenarkiv Vis sammendrag

  Most robot systems are based on a single decision-making process. This process needs to balance time, energy, and accuracy in every situation. However, according to ”dual process theory” (DPT) from cognitive psychology, this is not how humans work. Depending on the situation, we have the ability to switch between two thinking methods, a fast system 1 (S1) and a slower system 2 (S2). In this paper, we propose a novel approach to a dual process architecture for robots and agents. Our method, called Dual Process Dreamer (DPDreamer), is a combination of a reinforcement learning policy network, a planning algorithm, and a learned world model. The world model allows the parts of DPDreamer to work together and create a more integrated system compared to previous proposals of DPT systems. DPDreamer was tested in a puzzle game called Sokoban, and by balancing the use of S1 and S2, DPDreamer managed a success rate similar to S2 while using S1 most of the time, showing the benefit of using a more adaptable system.

• Lømo, Tobias; Tørresen, Jim; Kolberg, Mariana & Maffei, Renan (2024). Multi Map Visual Localization for Unmanned Aerial Vehicles. IEEE Robotics and Automation Letters. ISSN 2377-3766. 10(2). doi: 10.1109/LRA.2024.3518071. Fulltekst i vitenarkiv Vis sammendrag

  Localization has long been an essential area of research within robotics. The popularity of using Unmanned Aerial Vehicles (UAVs) to solve different tasks has increased and is expected to continue. Developing a robust complementary system to the Global Navigation Satellite Systems (GNSS) used today has been researched, and visual localization using cameras and satellite images is a popular choice to use. One of the challenges with using satellite images is that different images over the same area can impact the system's performance. This article proposes a novel approach called Multi Map Visual Localization (MMVL), a method to use multiple satellite images simultaneously, which is combined using a weighted average of probability maps. The proposal uses a convolutional neural network (CNN) with a caching strategy together with Monto Carlo Localization (MCL). MMVL achieves excellent robustness compared to other approaches and manages to estimate the correct location on all test flights. At the same time, using multiple satellite images does not significantly impact accuracy and computation time.

Se alle arbeider i NVA

• Lømo, Tobias; Maffei, Renan; Kolberg, Mariana & Tørresen, Jim (2024). Visual localization of unmanned aerial vehicles using multiple satellite maps. Universitetet i Oslo. Fulltekst i vitenarkiv

Se alle arbeider i NVA