Biology subjects

Belkaid, M.

Publications and source records attributed to Belkaid, M..

5 recordsLinked to original sources

bioRxiv2026

A promptable foundation model enables automated multi-task dataset construction and real-time pose estimation in mice

Behavioural experiments often hinge on moments when animals crowd together. In these situations, existing pose-estimation methods usually fail because detected crops include neighbouring animals, and their body parts are easily confused between individuals. We solved it with DeepLabSAM, a pipeline that segments and tracks every animal from a single text prompt using the SAM 3 foundation model, followed by a mask-gated pretrained pose estimation network from DeepLabCut. Across recordings of multiple interacting mice, the proposed method substantially reduces misassigned keypoints under heavy occlusion. Furthermore, we used DeepLabSAM to automatically construct the Multi-Task Mouse Behaviour (MTMB) dataset, comprising 14,300 frames spanning nine behavioural assays, with each instance annotated for bounding box, mask, identity, and pose. SuperAnimal pretrained pose models transferred unpredictably across this dataset. Conversely, MouseLite, our RF-DETR pose-estimation model trained on the MTMB dataset, matched their in-distribution accuracy, generalized better to unseen assays, and achieved real-time (faster-than-acquisition) inference. Combined, DeepLabSAM, MTMB dataset, and MouseLite provide a scalable framework for robust, cross-assay behavioural pose tracking of freely interacting mice.

animal behavior and cognition↗

bioRxiv2026

Using Disinhibition versus Direct Control in a Spiking Neural Model of Dopamine-Driven Reinforcement Learning

Dopaminergic signalling is central to value learning and decision making. It has been observed that multiple pathways with different patterns of connectivity project to midbrain dopaminergic neurons, some involving direct excitatory projections while others involve disinhibition. However, the respective contributions of these patterns to dopamine control, and their computational and functional advantages remain unclear. In the current work we simulate and evaluate two fully spiking neural models of dopaminergic control, based either solely on disinhibition, or solely on direct inhibitory and excitatory projections. We compare these models in terms of their engineering properties, their resulting spiking profiles, and their ability to successfully acquire representations of expected value in a 3-armed bandit task. We find that both models are able to operate at an asynchronous-irregular firing regime, but that the firing profile of the direct integration model is less resilient to disruption and more sensitive to incoming signals. In addition, the disinhibition model performs better in the learning task. We conclude that while the direct model is more parsimonious, disinhibition-based control remains advantageous in the operational context. Our results have implications for the study of decision-making brain circuits as well as for the design of brain-inspired systems.

neuroscience↗

bioRxiv2020

The cowboy effect: robot gaze influences human decisions

In most everyday life situations, the brain needs to engage not only in making decisions, but also in anticipating and predicting the behavior of others. In such contexts, gaze can be highly informative about others intentions, goals and upcoming decisions. Here, we investigated whether a humanoid robots gaze (mutual or averted) influences the way people strategically reason in a social decisionmaking context. Specifically, participants played a strategic game with the robot iCub while we measured their behavior and neural (EEG) activity. Participants were slower to respond when iCub established mutual gaze prior to their decision, relative to averted gaze. This was associated with a higher decision threshold in the drift diffusion model and accompanied by more synchronized EEG alpha activity. In addition, we found that participants reasoned about the robots actions in both conditions. However, those who mostly experienced the averted gaze were more likely to adopt a self-oriented strategy and their neural activity showed higher sensitivity to outcome. Altogether, these findings suggest that robot gaze acts as a strong social signal for humans, modulating response times, decision threshold, neural synchronization, as well as choice strategies and sensitivity outcomes. This has strong implications for all contexts involving human-robot interaction, from robotics to clinical applications.

neuroscience↗

bioRxiv2019

Modeling Uncertainty-Seeking Behavior Mediated by Cholinergic Influence on Dopamine

Recent findings suggest that acetylcholine mediates uncertainty-seeking behaviors through its projection to dopamine neurons - another neuromodulatory system known for its major implication in reinforcement learning and decision-making. In this paper, we propose a leaky-integrate-and-fire model of this mechanism. It implements a softmax-like selection with an uncertainty bonus by a cholinergic drive to dopaminergic neurons, which in turn influence synaptic currents of downstream neurons. The model is able to reproduce experimental data in two decision-making tasks. It also predicts that i) in the absence of cholinergic input, dopaminergic activity would not correlate with uncertainty, and that ii) the adaptive advantage brought by the implemented uncertainty-seeking mechanism is most useful when sources of reward are not highly uncertain. Moreover, this modeling work allows us to propose novel experiments which might shed new light on the role of acetylcholine in both random and directed exploration. Overall, this study thus contributes to a more comprehensive understanding of the roles of the cholinergic system and its involvement in decision-making in particular.

neuroscience↗

bioRxiv2019

Mice adaptively generate choice variability in a deterministic task

Can our choices just be driven by chance? To investigate this question, we designed a deterministic setting in which mice reinforce non-repetitive choice sequences, and modeled it using reinforcement learning. Mice progressively increased their choice variability using a memory-free, pseudo-random selection, rather than by learning complex sequences. Our results demonstrate that a decision-making process can self-generate variability and randomness even when the rules governing reward delivery are not stochastic.

neuroscience↗