arXiv subjects

David Colliaux

Publications and source records attributed to David Colliaux.

8 recordsLinked to original sources

arXiv2026

Active sensing to characterize the heterogeneity of plant stress

While most phenotyping platforms rely primarily on image-based measurements, advanced plant characterization requires the integration of active physiological sensing modali- ties such as chlorophyll fluorescence. We present an autonomous robotic platform designed to perform targeted fluorescence measurements on plant leaves. The system combines 3D plant reconstruction, geometric analysis, and motion planning to localize suitable measurement points and generate collision-free trajectories for a robotic manipulator. A dense 3D model of the plant is reconstructed from multi-view data and used to extract candidate leaf surfaces based on orientation, accessibility, and sensing constraints. These targets are then integrated into a task-level planning framework that guides the end-effector to precise contact or near-contact configurations required for point-based fluorescence acquisition. The platform enables automated, repeatable, and spatially resolved physiological measurements that go beyond passive imaging. By tightly coupling perception, geometric reasoning, and manipulation, the proposed system provides a robotics-driven approach to high-resolution plant phenotyping and opens new directions for autonomous agricultural inspection and plant-aware manipulation.

cs.RO↗

arXiv2026

Implementation of reinforcement learning in chemical reaction networks: application to phototaxis as curiosity-driven exploration

Living systems navigate environments using noisy and incomplete sensory signals. In unicellular algae, phototaxis is often modeled as a mechanistic run--tumble process driven by stimulus--response rules. However, such descriptions overlook how organisms actively sample their environment to reduce sensory ambiguity. From a minimal cognition perspective, we reframe this navigation as a subjective, information-driven sensorimotor process. To this end, we propose a framework linking a Partially Observable Markov Decision Process (POMDP) with biochemical reaction dynamics. Environmental variables are hidden, while the cell updates a minimal internal state from each observation through a memoryless Bayesian step. These internal dynamics balance orienting toward light with exploratory reorientation and can be implemented through Chemical-Reaction-Network Ordinary Differential Equations (CRN--ODEs). Our model includes a biophysical observation process for photoreception and a chemically computable polynomial bound on information gain. Using Inverse Reinforcement Learning (IRL) on 30 experimentally recorded Chlamydomonas trajectories, we infer the behavioral objective consistent with observed phototactic motion and benchmark the resulting dynamics with standard Stochastic Simulation Algorithm (SSA) baselines. Our model reproduces the empirical alignment-to-light distribution, comparable to objective SSA baselines on this dataset. Within this framework, run--tumble alternation emerges as an information-acquisition strategy: tumbling reorients the cell to sample new sensory configurations and resolve sensor ambiguity, demonstrating how intracellular biochemical networks can support adaptive information-seeking behavior in cellular navigation.

cs.LG↗

arXiv2026

Natural Gradient Descent for Online Continual Learning

Online Continual Learning (OCL) for image classification represents a challenging subset of Continual Learning, focusing on classifying images from a stream without assuming data independence and identical distribution (i.i.d). The primary challenge in this context is to prevent catastrophic forgetting, where the model's performance on previous tasks deteriorates as it learns new ones. Although various strategies have been proposed to address this issue, achieving rapid convergence remains a significant challenge in the online setting. In this work, we introduce a novel approach to training OCL models that utilizes the Natural Gradient Descent optimizer, incorporating an approximation of the Fisher Information Matrix (FIM) through Kronecker Factored Approximate Curvature (KFAC). This method demonstrates substantial improvements in performance across all OCL methods, particularly when combined with existing OCL tricks, on datasets such as Split CIFAR-100, CORE50, and Split miniImageNet.

cs.LG↗

arXiv2026

Altar: Structuring Sharable Experimental Data from Early Exploration to Publication

Managing the data and metadata during the active development phase of an experimental project presents a significant challenge, particularly in collaborative research. This phase is frequently overlooked in Data Management Plans included in project proposals, despite its important role in ensuring reproducibility and preventing the need for retroactive reconstruction at the time of publication. Here we present Altar, a lightweight, domain-agnostic framework for structuring experimental data from the onset of a project without imposing rigid data models. Altar is built around the Sacred experiment-tracking model and captures experimental (meta)data and structures them. Parameters, metadata, curves and small files are stored in a flexible NoSQL database, while large raw data are maintained in dedicated storage and linked through unique identifiers, ensuring efficiency and traceability. This integration is composable with exiting workflows, allowing integration with minimial disruption of work habits. We document different pathways to use Altar based on users skillset (PhD students, Post-docs, Principal Investigators, Laboratory administrators, System administrators). While getting started with Altar does not require a specialized infrastructure, the framework can be easily deployed on a server and made publicly accessible when scaling up or preparing data for publication. By addressing the dynamic phase of research, Altar provides a practical bridge between exploratory experimentation and FAIR-aligned data sharing.

cs.IR↗

arXiv2024

A random planting model

The adoption of agroecological practices will be crucial to address the challenges of climate change and biodiversity loss. Such practices favor the cultivation of plants in complex mixtures with layouts differing from the monoculture approach of conventional agriculture. Inspired by random sequential adsorption processes, we propose a one-dimensional model in which the plants are represented as line segments that start as points and grow at a constant rate until they reach length $\sigma$ after a time interval $\tau$. The planting positions and times are randomly chosen with the constraint that plant overlap is forbidden. We apply an exact, event-driven simulation to investigate the resulting spatiotemporal patterns and yields in both mono- and duocultures. After a transient period, with oscillations in the density and coverage, the field reaches a steady state in which the mean age of plants is one half of the time to maturity. The structure of the active plants is characterized by correlation functions between the fluctuation of the age of a plant and its $k$th neighbour. Nearest neighbours are negatively correlated, while next nearest neighbours tend to have similar ages. The steady state yield increases with the planting rate and approaches a maximum value of 4/3 plants per unit length per unit time. For two species with the same size at maturity but different growth rates, the more slowly growing species is enriched in the harvest compared to the seed mix composition. If two species have the same time to maturity but different sizes, the smaller one is enriched in the harvest and, at a sufficiently high planting rate, the larger species may be completely absent. For two species with the same ratio of $\sigma/\tau$ the selectivity is insensitive to the planting rate. The model may be extended to higher dimensions, more species and other planting strategies.

q-bio.PE↗

arXiv2023

Continual Learning with Deep Streaming Regularized Discriminant Analysis

Continual learning is increasingly sought after in real world machine learning applications, as it enables learning in a more human-like manner. Conventional machine learning approaches fail to achieve this, as incrementally updating the model with non-identically distributed data leads to catastrophic forgetting, where existing representations are overwritten. Although traditional continual learning methods have mostly focused on batch learning, which involves learning from large collections of labeled data sequentially, this approach is not well-suited for real-world applications where we would like new data to be integrated directly. This necessitates a paradigm shift towards streaming learning. In this paper, we propose a streaming version of regularized discriminant analysis as a solution to this challenge. We combine our algorithm with a convolutional neural network and demonstrate that it outperforms both batch learning and existing streaming learning algorithms on the ImageNet ILSVRC-2012 dataset.

cs.CV↗

arXiv2020

Intrinsic Motivation and Episodic Memories for Robot Exploration of High-Dimensional Sensory Spaces

This work presents an architecture that generates curiosity-driven goal-directed exploration behaviours for an image sensor of a microfarming robot. A combination of deep neural networks for offline unsupervised learning of low-dimensional features from images, and of online learning of shallow neural networks representing the inverse and forward kinematics of the system have been used. The artificial curiosity system assigns interest values to a set of pre-defined goals, and drives the exploration towards those that are expected to maximise the learning progress. We propose the integration of an episodic memory in intrinsic motivation systems to face catastrophic forgetting issues, typically experienced when performing online updates of artificial neural networks. Our results show that adopting an episodic memory system not only prevents the computational models from quickly forgetting knowledge that has been previously acquired, but also provides new avenues for modulating the balance between plasticity and stability of the models.

cs.AI↗

arXiv2015

Toward Biochemical Probabilistic Computation

Living organisms survive and multiply even though they have uncertain and incomplete information about their environment and imperfect models to predict the consequences of their actions. Bayesian models have been proposed to face this challenge. Indeed, Bayesian inference is a way to do optimal reasoning when only uncertain and incomplete information is available. Various perceptive, sensory-motor, and cognitive functions have been successfully modeled this way. However, the biological mechanisms allowing animals and humans to represent and to compute probability distributions are not known. It has been proposed that neurons and assemblies of neurons could be the appropriate scale to search for clues to probabilistic reasoning. In contrast, in this paper, we propose that interacting populations of macromolecules and diffusible messengers can perform probabilistic computation. This suggests that probabilistic reasoning, based on cellular signaling pathways, is a fundamental skill of living organisms available to the simplest unicellular organisms as well as the most complex brains.

cs.ET↗