A unified probabilistic model of the perception of three-dimensional structure from optic flow

Don't worry, be active: how to facilitate the detection of errors in immersive virtual environments

The current research aims to study the link between the type of vision experienced in a collaborative immersive virtual environment (active vs. multiple passive), the type of error one looks for during a cooperative multi-user exploration of a design project (affordance vs. perceptual violations), and the type of setting in which multi-user perform (field in Experiment 1 vs. laboratory in Experiment 2). The relevance of this link is backed by the lack of conclusive evidence on an active vs. passive vision advantage in cooperative search tasks within software based on immersive virtual reality (IVR). Using a yoking paradigm based on the mixed usage of simultaneous active and multiple passive viewings, we found that the likelihood of error detection in a complex 3D environment was characterized by an active vs. multi-passive viewing advantage depending on: (1) the degree of knowledge dependence of the type of error the passive/active observers were looking for (low for perceptual violations, vs. high for affordance violations), as the advantage tended to manifest itself irrespectively from the setting for affordance, but not for perceptual violations; and (2) the degree of social desirability possibly induced by the setting in which the task was performed, as the advantage occurred irrespectively from the type of error in the laboratory (Experiment 2) but not in the field (Experiment 1) setting. Results are relevant to future development of cooperative software based on IVR used for supporting the design review. A multi-user design review experience in which designers, engineers and end-users all cooperate actively within the IVR wearing their own head mounted display, seems more suitable for the detection of relevant errors than standard systems characterized by a mixed usage of active and passive viewing.

Illusory Tactile Motion Perception: An Analog of the Visual Filehne Illusion

We continually move our body and our eyes when exploring the world, causing our sensory surfaces, the skin and the retina, to move relative to external objects. In order to estimate object motion consistently, an ideal observer would transform estimates of motion acquired from the sensory surface into fixed, world-centered estimates, by taking the motion of the sensor into account. This ability is referred to as spatial constancy. Human vision does not follow this rule strictly and is therefore subject to perceptual illusions during eye movements, where immobile objects can appear to move. Here, we investigated whether one of these, the Filehne illusion, had a counterpart in touch. To this end, observers estimated the movement of a surface from tactile slip, with a moving or with a stationary finger. We found the perceived movement of the surface to be biased if the surface was sensed while moving. This effect exemplifies a failure of spatial constancy that is similar to the Filehne illusion in vision. We quantified this illusion by using a Bayesian model with a prior for stationarity, applied previously in vision. The analogy between vision and touch points to a modality-independent solution to the spatial constancy problem.

Perceived surface slant is systematically biased in the actively-generated optic flow

Humans make systematic errors in the 3D interpretation of the optic flow in both passive and active vision. These systematic distortions can be predicted by a biologically-inspired model which disregards self-motion information resulting from head movements (Caudek, Fantoni, & Domini 2011). Here, we tested two predictions of this model: (1) A plane that is stationary in an earth-fixed reference frame will be perceived as changing its slant if the movement of the observer's head causes a variation of the optic flow; (2) a surface that rotates in an earth-fixed reference frame will be perceived to be stationary, if the surface rotation is appropriately yoked to the head movement so as to generate a variation of the surface slant but not of the optic flow. Both predictions were corroborated by two experiments in which observers judged the perceived slant of a random-dot planar surface during egomotion. We found qualitatively similar biases for monocular and binocular viewing of the simulated surfaces, although, in principle, the simultaneous presence of disparity and motion cues allows for a veridical recovery of surface slant.

Bayesian action-perception computational model: interaction of production and recognition of cursive letters

In this paper, we study the collaboration of perception and action representations involved in cursive letter recognition and production. We propose a mathematical formulation for the whole perception-action loop, based on probabilistic modeling and bayesian inference, which we call the Bayesian Action-Perception (BAP) model. Being a model of both perception and action processes, the purpose of this model is to study the interaction of these processes. More precisely, the model includes a feedback loop from motor production, which implements an internal simulation of movement. Motor knowledge can therefore be involved during perception tasks. In this paper, we formally define the BAP model and show how it solves the following six varied cognitive tasks using bayesian inference: i) letter recognition (purely sensory), ii) writer recognition, iii) letter production (with different effectors), iv) copying of trajectories, v) copying of letters, and vi) letter recognition (with internal simulation of movements). We present computer simulations of each of these cognitive tasks, and discuss experimental predictions and theoretical developments.

Bayesian modeling of perceived surface slant from actively-generated and passively-observed optic flow

We measured perceived depth from the optic flow (a) when showing a stationary physical or virtual object to observers who moved their head at a normal or slower speed, and (b) when simulating the same optic flow on a computer and presenting it to stationary observers. Our results show that perceived surface slant is systematically distorted, for both the active and the passive viewing of physical or virtual surfaces. These distortions are modulated by head translation speed, with perceived slant increasing directly with the local velocity gradient of the optic flow. This empirical result allows us to determine the relative merits of two alternative approaches aimed at explaining perceived surface slant in active vision: an "inverse optics" model that takes head motion information into account, and a probabilistic model that ignores extra-retinal signals. We compare these two approaches within the framework of the bayesian theory. The "inverse optics" bayesian model produces veridical slant estimates if the optic flow and the head translation velocity are measured with no error; because of the influence of a "prior" for flatness, the slant estimates become systematically biased as the measurement errors increase. The bayesian model, which ignores the observer's motion, always produces distorted estimates of surface slant. Interestingly, the predictions of this second model, not those of the first one, are consistent with our empirical findings. The present results suggest that (a) in active vision perceived surface slant may be the product of probabilistic processes which do not guarantee the correct solution, and (b) extra-retinal signals may be mainly used for a better measurement of retinal information.

The probabilistic cell: implementation of a probabilistic inference by the biochemical mechanisms of phototransduction

When we perceive the external world, our brain has to deal with the incompleteness and uncertainty associated with sensory inputs, memory and prior knowledge. In theoretical neuroscience probabilistic approaches have received a growing interest recently, as they account for the ability to reason with incomplete knowledge and to efficiently describe perceptive and behavioral tasks. How can the probability distributions that need to be estimated in these models be represented and processed in the brain, in particular at the single cell level? We consider the basic function carried out by photoreceptor cells which consists in detecting the presence or absence of light. We give a system-level understanding of the process of phototransduction based on a bayesian formalism: we show that the process of phototransduction is equivalent to a temporal probabilistic inference in a Hidden Markov Model (HMM), for estimating the presence or absence of light. Thus, the biochemical mechanisms of phototransduction underlie the estimation of the current state probability distribution of the presence of light. A classical descriptive model describes the interactions between the different molecular messengers, ions, enzymes and channel proteins occurring within the photoreceptor by a set of nonlinear coupled differential equations. In contrast, the probabilistic HMM model is described by a discrete recurrence equation. It appears that the binary HMM has a general solution in the case of constant input. This allows a detailed analysis of the dynamics of the system. The biochemical system and the HMM behave similarly under steady-state conditions. Consequently a formal equivalence can be found between the biochemical system and the HMM. Numerical simulations further extend the results to the dynamic case and to noisy input. All in all, we have derived a probabilistic model equivalent to a classical descriptive model of phototransduction, which has the additional advantage of assigning a function to phototransduction. The example of phototransduction shows how simple biochemical interactions underlie simple probabilistic inferences.

Observer Movement and Size Constancy

It is commonly assumed that size constancy—invariance of perceived size of objects as they change retinal size because of changes in distance—depends solely on retinal stimulation and vergence, but on no other action-related signals. Distance to an object can change through displacement of either the observer or the object. The common assumption predicts that the two types of displacement should lead to the same degree of size constancy. We measured size constancy while observers viewed stationary stimuli at different distances. Changes in distance between trials were either actively produced by the observer or generated by real or simulated object displacement, with retinal stimulation held constant across the movement conditions. Responses were always closer to perfect constancy for observer than for object movement. Thus, size constancy is enhanced by information from observer displacement, and, more generally, processes thought to be purely perceptual may have unexpected components related to action.