David Filliat
YOU?
Author Swipe
View article: Interpretable Rule Learning for Reactive Activity Recognition in Event-Driven RL Environments
Interpretable Rule Learning for Reactive Activity Recognition in Event-Driven RL Environments Open
International audience
View article: NAMOUnc: Navigation Among Movable Obstacles with Decision Making on Uncertainty Interval
NAMOUnc: Navigation Among Movable Obstacles with Decision Making on Uncertainty Interval Open
Navigation among movable obstacles (NAMO) is a critical task in robotics, often challenged by real-world uncertainties such as observation noise, model approximations, action failures, and partial observability. Existing solutions frequent…
View article: Improved monocular depth prediction using distance transform over pre-semantic contours with self-supervised neural networks
Improved monocular depth prediction using distance transform over pre-semantic contours with self-supervised neural networks Open
International audience
View article: A Simple yet Effective Test-Time Adaptation for Zero-Shot Monocular Metric Depth Estimation
A Simple yet Effective Test-Time Adaptation for Zero-Shot Monocular Metric Depth Estimation Open
The recent development of \emph{foundation models} for monocular depth estimation such as Depth Anything paved the way to zero-shot monocular depth estimation. Since it returns an affine-invariant disparity map, the favored technique to re…
View article: Hierarchical Light Transformer Ensembles for Multimodal Trajectory Forecasting
Hierarchical Light Transformer Ensembles for Multimodal Trajectory Forecasting Open
Accurate trajectory forecasting is crucial for the performance of various systems, such as advanced driver-assistance systems and self-driving vehicles. These forecasts allow us to anticipate events that lead to collisions and, therefore, …
View article: InfraParis: A multi-modal and multi-task autonomous driving dataset
InfraParis: A multi-modal and multi-task autonomous driving dataset Open
Current deep neural networks (DNNs) for autonomous driving computer vision are typically trained on specific datasets that only involve a single type of data and urban scenes. Consequently, these models struggle to handle new objects, nois…
View article: Open the Chests: An Environment for Activity Recognition and Sequential Decision Problems Using Temporal Logic
Open the Chests: An Environment for Activity Recognition and Sequential Decision Problems Using Temporal Logic Open
This article presents Open the Chests, a novel benchmark environment designed for simulating and testing activity recognition and reactive decision-making algorithms. By leveraging temporal logic, Open the Chests offers a dynamic, event-dr…
View article: On Double Descent in Reinforcement Learning with LSTD and Random Features
On Double Descent in Reinforcement Learning with LSTD and Random Features Open
Temporal Difference (TD) algorithms are widely used in Deep Reinforcement Learning (RL). Their performance is heavily influenced by the size of the neural network. While in supervised learning, the regime of over-parameterization and its b…
View article: Fourier Features in Reinforcement Learning with Neural Networks
Fourier Features in Reinforcement Learning with Neural Networks Open
International audience
View article: VIBR: Learning View-Invariant Value Functions for Robust Visual Control
VIBR: Learning View-Invariant Value Functions for Robust Visual Control Open
End-to-end reinforcement learning on images showed significant progress in the recent years. Data-based approach leverage data augmentation and domain randomization while representation learning methods use auxiliary losses to learn task-r…
View article: Domain invariant Q-learning for model-free robust continuous control under visual distractions
Domain invariant Q-learning for model-free robust continuous control under visual distractions Open
International audience
View article: MUAD: Multiple Uncertainties for Autonomous Driving
MUAD: Multiple Uncertainties for Autonomous Driving Open
We introduce MUAD, a synthetic dataset for autonomous driving with multiple uncertainty types and tasks. It contains 10413 in total: 3420 images in the train set, 492 in the validation set and 6501 in the test set. The test set is divided …
View article: Latent Discriminant deterministic Uncertainty
Latent Discriminant deterministic Uncertainty Open
Predictive uncertainty estimation is essential for deploying Deep Neural Networks in real-world autonomous systems. However, most successful approaches are computationally intensive. In this work, we attempt to address these challenges in …
View article: A study of deep perceptual metrics for image quality assessment
A study of deep perceptual metrics for image quality assessment Open
Several metrics exist to quantify the similarity between images, but they are inefficient when it comes to measure the similarity of highly distorted images. In this work, we propose to empirically investigate perceptual metrics based on d…
View article: Task and Motion Planning Methods: Applications and Limitations
Task and Motion Planning Methods: Applications and Limitations Open
International audience
View article: EXplainable Neural-Symbolic Learning (X-NeSyL) methodology to fuse deep learning representations with expert knowledge graphs: The MonuMAI cultural heritage use case
EXplainable Neural-Symbolic Learning (X-NeSyL) methodology to fuse deep learning representations with expert knowledge graphs: The MonuMAI cultural heritage use case Open
The latest Deep Learning (DL) models for detection and classification have achieved an unprecedented performance over classical machine learning algorithms. However, DL models are black-box methods hard to debug, interpret, and certify. DL…
View article: POAR: Efficient Policy Optimization via Online Abstract State Representation Learning
POAR: Efficient Policy Optimization via Online Abstract State Representation Learning Open
While the rapid progress of deep learning fuels end-to-end reinforcement learning (RL), direct application, especially in high-dimensional space like robotic scenarios still suffers from low sample efficiency. Therefore State Representatio…
View article: S-TRIGGER: Continual State Representation Learning via Self-Triggered Generative Replay
S-TRIGGER: Continual State Representation Learning via Self-Triggered Generative Replay Open
We consider the problem of building a state representation model for control, in a continual learning setting. As the environment changes, the aim is to efficiently compress the sensory state's information without losing past knowledge, an…
View article: Evaluating Robustness over High Level Driving Instruction for Autonomous Driving
Evaluating Robustness over High Level Driving Instruction for Autonomous Driving Open
In recent years, we have witnessed increasingly high performance in the field of autonomous end-to-end driving. In particular, more and more research is being done on driving in urban environments, where the car has to follow high level co…
View article: SCOD: Active Object Detection for Embodied Agents using Sensory Commutativity of Action Sequences
SCOD: Active Object Detection for Embodied Agents using Sensory Commutativity of Action Sequences Open
We introduce SCOD (Sensory Commutativity Object Detection), an active method for movable and immovable object detection. SCOD exploits the commutative properties of action sequences, in the scenario of an embodied agent equipped with first…
View article: Are standard Object Segmentation models sufficient for Learning Affordance Segmentation?
Are standard Object Segmentation models sufficient for Learning Affordance Segmentation? Open
Affordances are the possibilities of actions the environment offers to the individual. Ordinary objects (hammer, knife) usually have many affordances (grasping, pounding, cutting), and detecting these allow artificial agents to understand …
View article: Evaluating Robustness over High Level Driving Instruction for Autonomous\n Driving
Evaluating Robustness over High Level Driving Instruction for Autonomous\n Driving Open
In recent years, we have witnessed increasingly high performance in the field\nof autonomous end-to-end driving. In particular, more and more research is\nbeing done on driving in urban environments, where the car has to follow high\nlevel…
View article: On the Sensory Commutativity of Action Sequences for Embodied Agents
On the Sensory Commutativity of Action Sequences for Embodied Agents Open
Perception of artificial agents is one the grand challenges of AI research. Deep Learning and data-driven approaches are successful on constrained problems where perception can be learned using supervision, but do not scale to open-worlds.…