Tom De Schepper

• Doctor in Science: Computer Science
• Principle Member of Technical Staff at imec

Learning to communicate in order to share state information is an active problem in the area of multi-agent reinforcement learning. The credit assignment problem, the non-stationarity of the communication environment and the problem of encouraging the agents to be influenced by incoming messages are major challenges within this research field which...

Multimodal networks have demonstrated remarkable performance improvements over their unimodal counterparts. Existing multimodal networks are designed in a multi-branch fashion that, due to the reliance on fusion strategies, exhibit deteriorated performance if one or more modalities are missing. In this work, we propose a modality invariant multimod...

Interest in deploying deep reinforcement learning (DRL) models on low-power edge devices, such as Autonomous Mobile Robots (AMRs) and Internet of Things (IoT) devices, has seen a significant rise due to the potential of performing real-time inference by eliminating the latency and reliability issues incurred from wireless communication and the priv...

Multimodal learning has demonstrated remarkable performance improvements over unimodal architectures. However, multimodal learning methods often exhibit deteriorated performances if one or more modalities are missing. This may be attributed to the commonly used multi-branch design containing modality-specific streams making the models reliant on th...

Dataset condensation (DC) methods aim to learn a smaller synthesized dataset with informative data records to accelerate the training of machine learning models. Current distribution matching (DM) based DC methods learn a synthesized dataset by matching the mean of the latent embeddings between the synthetic and the real dataset. However two distri...

Changes in climate can greatly affect the phenology of plants, which can have important feedback effects, such as altering the carbon cycle. These phenologi-cal feedback effects are often induced by a shift in the start or end dates of the growing season of plants. The normalized difference vegetation index (NDVI) serves as a straightforward indica...

Communication is crucial in multi-agent reinforcement learning when agents are not able to observe the full state of the environment. The most common approach to allow learned communication between agents is the use of a differentiable communication channel that allows gradients to flow between agents as a form of feedback. However, this is challen...

Many multi-agent systems require inter-agent communication to properly achieve their goal. By learning the communication protocol alongside the action protocol using multi-agent reinforcement learning techniques, the agents gain the flexibility to determine which information should be shared. However, when the number of agents increases we need to...

We define Location Category Inference (LCI) as a task of predicting the category of a visited venue, such as bar, restaurant or university, given user location GPS coordinates and a set of venue candidates. LCI is an essential part of the hyper-personalization systems as its output provides deep insights into user lifestyle (has children, owns a do...

In recent years multi-label, multi-class video action recognition has gained significant popularity. While reasoning over temporally connected atomic actions is mundane for intelligent species, standard artificial neural networks (ANN) still struggle to classify them. In the real world, atomic actions often temporally connect to form more complex c...

Recent advancements have made Deep Reinforcement Learning (DRL) exceedingly more powerful, but the produced models remain very computationally complex and therefore difficult to deploy on edge devices. Compression methods such as quantization and distillation can be used to increase the applicability of DRL models on these low-power edge devices by...

Finding an object of a specific class in an unseen environment remains an unsolved navigation problem. Hence, we propose a hierarchical learning-based method for object navigation. The top-level is capable of high-level planning, and building a memory on a floorplan-level (e.g., which room makes the most sense for the agent to visit next, where has...

Communication is crucial in multi-agent reinforcement learning when agents are not able to observe the full state of the environment. The most common approach to allow learned communication between agents is the use of a differentiable communication channel that allows gradients to flow between agents as a form of feedback. However, this is challen...

In the paper, we describe the technical details of a multi-player tracker system using tracking data obtained from a single low-cost stationary camera on field hockey games. Analyzing the tracking data of the players only from the transmitted video opens a multitude of applications that allows the cost of technology to be reduced. This method does...

Object detection on real-time edge devices for new applications with no or a limited amount of annotated labels is difficult. Where traditional data-hungry methods fail, transfer learning can provide a solution by transferring knowledge from a source domain to the target application domain. We explore domain adaptation techniques on a one-stage det...

In recent years, Capsule Networks (CapsNets) have achieved promising results in tasks such as object recognition thanks to their invariance characteristics towards pose and lighting. They have been proposed as an alternative to relational insensitive and translation invariant Convolutional Neural Networks (CNN). It has been empirically proven that...

In recent years multi-label, multi-class video action recognition has gained significant popularity. While reasoning over temporally connected atomic actions is mundane for intelligent species, standard artificial neural networks (ANN) still struggle to classify them. In the real world, atomic actions often temporally connect to form more complex c...

Professional road cycling is a very competitive sport, and many factors influence the outcome of the race. These factors can be internal (e.g., psychological preparedness, physiological profile of the rider, and the preparedness or fitness of the rider) or external (e.g., the weather or strategy of the team) to the rider, or even completely unpredi...

Embodied AI, learning through interaction with a physical environment, typically requires large amounts of interaction with the environment in order to learn how to solve new tasks. Training can be done in parallel, using simulated environments. However, once deployed in e.g., a real-world setting, it is not yet clear how an agent can quickly adapt...

Predicting cycling race results has always been a task left to experts with a lot of domain knowledge. This is largely due to the fact that the outcomes of cycling races can be rather surprising and depend on an extensive set of parameters. Examples of such factors are, among others, the preparedness of a rider, the weather, the team strategy, and...

Today’s and tomorrow’s networks are becoming increasingly complex and heterogeneous with a large diversity of devices and technologies. To meet growing demand, and support client mobility there is need for intelligent mechanisms like multi-technology load balancing and handovers. Current solutions, like Multipath Transmission Control Protocol (MPTC...

Traffic recognition is commonly done using deep packet inspection or packet‐based approaches. However, these methods require the listening device to be part of the network and raise privacy concerns. Traffic recognition models that operate directly at the spectrum level could, for instance, be used for smart spectrum management. To this extent, we...

The number of connected devices has reached 18 billion in 2017 and this will nearly double by 2022, while also new wireless communication technologies become available. Since these modern devices support the use of multiple communication technologies, efforts have been made to enable simultaneous usage and handovers between the different technologi...

Wireless devices have a plethora of technologies at their disposal to connect to the Internet and other services. Management and control of each technology are traditionally isolated, and coordination between technologies is nearly non-existent. This isolation leads to poor resource usage, which in turn reduces performance and service guarantees. T...

Over the last decades, we have witnessed a tremendous increase in the utilization and availability of wireless networks and devices. This growth is largely founded by the introduction of mobile devices and the Internet of Things paradigm. In parallel, the demands and expectations of users, in terms of connectivity, bandwidth, quality, and services,...

The utilization and size of today's wireless networks is continuously increasing, as more and more wireless communication technologies and connected devices are being added. As the use of multiple communication technologies is supported by modern devices, efforts have been made to allow these devices to utilize simultaneously and handover between d...

Modern connected devices are equipped with the ability to connect to the Internet using a variety of different wireless network technologies. Current network management solutions fail to provide a fine-grained, coordinated, and transparent answer to this heterogeneity, while the lower layers of the OSI stack simply ignore it by providing full separ...

Today’s electronic devices have multiple communication technologies available at any time. Currently, the application layer or the user needs to manually switch between them, depending on the networks in range. No holistic and adaptive approach exists that can manage all technologies and devices at once. To this extent, we previously proposed the i...

Today’s local area networks (LANs) consist of an ever-expanding number of heterogeneous consumer devices and communication technologies. Despite supporting multiple technologies, those devices tend to connect to the Internet using a single technology, based on predefined priorities. This static behavior does not allow the network to unlock its full...

Local area networks (LANs) are employed by a plethora of heterogeneous consumer devices, equipped with the ability to connect to the Internet using a variety of different wireless network technologies. Existing solutions and the lower layers of the OSI stack are unfit to cope with this heterogeneity. For instance, dynamical inter-technology switchi...

Wi-Fi network roaming is the act of moving a wireless device from one Wi-Fi access point (AP) to another Wi-Fi AP. In urban environments, where APs are densely deployed, users would greatly benefit from roaming between these APs. Standards for Wi-Fi-network roaming have been developed (e.g. IEEE 802.11r), but are rarely implemented. The absence of...

Current mobile consumer devices are equipped with the ability to connect to the Internet using a variety of heterogeneous wireless network technologies (e.g., Wi-Fi and LTE). These devices generally opt to statically connect using a single technology, based on predefined priorities. This static behavior does not allow the network to unlock its full...

We have seen an ever-expanding set of location-aware services and devices become broadly available over the past years. Despite these promising applications and thorough research, indoor localization remains a very challenging topic and (near) perfect accuracy continues to be an open research challenge. In this paper, we acknowledge the growth and...

The current local area networks (LANs) are occupied by a large variety of heterogeneous consumer devices, equipped with the ability to connect to the Internet using a variety of different network technologies (e.g., Ethernet, 2.4 and 5GHz Wi-Fi). Nevertheless, devices generally opt to statically connect using a single technology, based on predefine...

Today's local area networks (LANs) consist of a plethora of heterogeneous consumer devices, equipped with the ability to connect to the Internet using a variety of different network technologies (e.g., Ethernet, Power-Line, 2.4 and 5GHz Wi-Fi). Nevertheless, devices generally opt to statically connect using a single technology, based on predefined...

The popularity and availability of Head Mounted Displays (HMDs) has known an increase over the past few years. In general those devices need to be connected to a computer to function properly. Moreover, applications that project images on such an HMD respond typically only to user inputs like mouse and keyboard actions and head movement. In this pa...

In this thesis we propose a mobile framework that is capable of tracking the movements of users and translating them into actions in a virtual world. This concept will be demonstrated through the creation of a multiplayer first person shooter game. The players will be wearing an Oculus Rift, an Head Mounted Display that allows a user to view a virt...