How multisensory neurons solve causal inference

Sitting in a static railway carriage can produce illusory self-motion if the train on an adjoining track moves off. While our visual system registers motion, vestibular signals indicate that we are stationary. The brain is faced with a difficult challenge: is there a single cause of sensations (I am moving) or two causes (I am static, another train is moving)? If a single cause, integrating signals produces a more precise estimate of self-motion, but if not, one cue should be ignored. In many cases, this process of causal inference works without error, but how does the brain achieve it? Electrophysiological recordings show that the macaque medial superior temporal area contains many neurons that encode combinations of vestibular and visual motion cues. Some respond best to vestibular and visual motion in the same direction (“congruent” neurons), while others prefer opposing directions (“opposite” neurons). Congruent neurons could underlie cue integration, but the function of opposite neurons remains a puzzle. Here, we seek to explain this computational arrangement by training a neural network model to solve causal inference for motion estimation. Like biological systems, the model develops congruent and opposite units and recapitulates known behavioral and neurophysiological observations. We show that all units (both congruent and opposite) contribute to motion estimation. Importantly, however, it is the balance between their activity that distinguishes whether visual and vestibular cues should be integrated or separated. This explains the computational purpose of puzzling neural representations and shows how a relatively simple feedforward network can solve causal inference.

When my avatar’s movements make me feel I am moving: From natural-like stimuli to fully artificial ones in virtual reality

In virtual reality, users do not receive any visual information coming from their own body. Thus, avatars are often used, and they can be embodied which alters the body representation. We suggested that the perception of one’s own movements (i.e., kinaesthesia) can be altered as well. We investigated whether visual cues coming from an avatar can be used for kinaesthesia and to what extent such cues can deviate from natural ones. We used a paradigm in which the participant’s left forearm was moved passively, correlated with the movement of both forearms of the avatar. Such visuo-proprioceptive combination induces kinaesthetic illusions in the participant’s right forearm. The impact of the morphological similarity (semantic congruency) and of the visual perspective of the avatar (spatial congruency) was investigated. Results have indicated that avatar’s movements are processed as one’s own movements. Morphological similarity and first-person perspective were not necessary, but they reinforced the illusions. Thus, visual motion cues can strongly deviate from natural ones in morphology and perspective and still contribute to kinaesthesia.

Behavioral and neuronal underpinnings of safety in numbers in fruit flies

Living in a group allows individuals to decrease their defenses enabling other beneficial behaviors such as foraging. The detection of a threat through social cues is widely reported, however the safety cues that guide animals to break away from a defensive behavior and resume alternate activities remain elusive. Here we show that fruit flies displayed a graded decrease in freezing behavior, triggered by an inescapable threat, with increasing group sizes. Furthermore, flies used the cessation of movement of other flies as a cue of threat and its resumption as a cue of safety. Finally, we found that lobula columnar neurons, LC11, mediate the propensity for freezing flies to resume moving in response to the movement of others. By identifying visual motion cues, and the neurons involved in their processing, as the basis of a social safety cue this study brings new insights into the neuronal basis of safety in numbers.

Neural circuits mediating visual stabilization during active motion in zebrafish

ABSTRACTVisual stabilization is an inevitable requirement for animals during active motion interaction with the environment. Visual motion cues of the surroundings or induced by self-generated behaviors are perceived then trigger proper motor responses mediated by neural representations conceptualized as the internal model: one part of it predicts the consequences of sensory dynamics as a forward model, another part generates proper motor control as a reverse model. However, the neural circuits between the two models remain mostly unknown. Here, we demonstrate that an internal component, the efference copy, coordinated the two models in a push-pull manner by generating extra reset saccades during active motion processing in larval zebrafish. Calcium imaging indicated that the saccade preparation circuit is enhanced while the velocity integration circuit is inhibited during the interaction, balancing the internal representations from both directions. This is the first model of efference copy on visual stabilization beyond the sensorimotor stage.