Amodal Perception Research Articles

AmodalSynthDrive: A Synthetic Amodal Perception Dataset for Autonomous Driving

AmodalSynthDrive: A Synthetic Amodal Perception Dataset for Autonomous Driving

Generative Adversarial Network for Overcoming Occlusion in Images: A Survey

Although current computer vision systems are closer to the human intelligence when it comes to comprehending the visible world than previously, their performance is hindered when objects are partially occluded. Since we live in a dynamic and complex environment, we encounter more occluded objects than fully visible ones. Therefore, instilling the capability of amodal perception into those vision systems is crucial. However, overcoming occlusion is difficult and comes with its own challenges. The generative adversarial network (GAN), on the other hand, is renowned for its generative power in producing data from a random noise distribution that approaches the samples that come from real data distributions. In this survey, we outline the existing works wherein GAN is utilized in addressing the challenges of overcoming occlusion, namely amodal segmentation, amodal content completion, order recovery, and acquiring training data. We provide a summary of the type of GAN, loss function, the dataset, and the results of each work. We present an overview of the implemented GAN architectures in various applications of amodal completion. We also discuss the common objective functions that are applied in training GAN for occlusion-handling tasks. Lastly, we discuss several open issues and potential future directions.

Cognitive A Priori and Amodal Perception in the Focus of Experimental Philosophy

The article discusses the prospects of neurophenomenology as a substantiative theory with respect to the results of neurocognitive research. Some relevant ideas and theoretical findings of the phenomenology of Edmund Husserl, which are of great importance for the interpretation of experimental data from neuroscience, are consistently presented and analyzed. In particular, the emphasis is made on the procedure of analogizing apperception (appresentation), based on an even deeper pairing mechanism. In so doing, I consider and trace the evolution of these ideas in Husserl's works of different years and periods. As an example, that clearly demonstrates the universal a priori rock-bottom role of appresentation in cognition and perception, the procedure of amodal completion is chosen. Amodal completion is the process of perception by which an object is apprehended as a whole while some parts of it are occluded by other objects. These research seems to open up wide opportunities for neurophenomenology as the theoretical basis of the sciences of consciousness, mind and brain.

Motion behind occluder: Amodal perception and visual motion extrapolation

ABSTRACT One of the main challenges that our visual system must resolve is to judge the current position of an occluded moving object. This computation is known as motion extrapolation (ME), and it is required to perform several tasks: for example, judging when an occluded moving target collides with a given cue or whether it reappeared in time. These tasks may easily be encountered in more ecological settings, highlighting the importance of ME in our daily life. The goal of the current review is to synthesize the existing literature that investigated targets during occluded motion, in order to provide a general overview of the topic. To this aim, different aspects are discussed: occluded motion as a true perceptual phenomenon, the commonly used tasks in the literature, how perceptual and cognitive factors may affect individual performance, which mechanisms are supposed to be involved during ME and, finally, the neural correlates of occluded motion perception.

Amodal Segmentation Based on Visible Region Segmentation and Shape Prior

Almost all existing amodal segmentation methods make the inferences of occluded regions by using features corresponding to the whole image. This is against the human's amodal perception, where human uses the visible part and the shape prior knowledge of the target to infer the occluded region. To mimic the behavior of human and solve the ambiguity in the learning, we propose a framework, it firstly estimates a coarse visible mask and a coarse amodal mask. Then based on the coarse prediction, our model infers the amodal mask by concentrating on the visible region and utilizing the shape prior in the memory. In this way, features corresponding to background and occlusion can be suppressed for amodal mask estimation. Consequently, the amodal mask would not be affected by what the occlusion is given the same visible regions. The leverage of shape prior makes the amodal mask estimation more robust and reasonable. Our proposed model is evaluated on three datasets. Experiments show that our proposed model outperforms existing state-of-the-art methods. The visualization of shape prior indicates that the category-specific feature in the codebook has certain interpretability. The code is available at https://github.com/YutingXiao/Amodal-Segmentation-Based-on-Visible-Region-Segmentation-and-Shape-Prior.

News From Home the redux version: Amodal perception and ‘la jouissance du voir’

In my 2004 article on Chantal Akerman’s News From Home (1976), I deployed a Deleuzian reading of the film to release the film from being interpreted as manifesting the impossibility of a woman’s desire (Stephen Heath) or a desire to return to the mother (Richard Kwietniowski), arguing instead that the indeterminacy of the final sequence opens out onto a transformative freedom from identity. In this article, I persist with this idea but reconsider it in relation to Griselda Pollock’s convincing insistence that Akerman’s work is a journey towards maternal trauma, a position that she develops in relation to Akerman’s installation Walking Next to One’s Shoelaces Inside an Empty Fridge. Before encountering this work, Pollock says that Akerman’s cinematic intervention was linked to ‘the choked feminine voice in culture meeting a new cinematic formalism’ rather than to the ‘deeper trauma’ of being the child of a Holocaust survivor. Pollock’s convincing reading of Akerman’s installation as visualizing the effects of unmourned trauma transmitted to the children of Holocaust survivors had a profound impact on me, one that I take into consideration in my reframing in this article of the final sequence of News From Home. In this, I deploy Raymond Bellour’s adaptation of psychoanalyst Daniel Stern’s notion of ‘amodal perception’ as a sensory, kinetic modality of spectatorship. This model allows me to retain the transformative freedom from identity in my earlier reading, while nonetheless mapping early intersubjective relations onto the pleasures of the film body Akerman called ‘la jouissance du voir’.

There’s more to “sparkle” than meets the eye: Knowledge of vision and light verbs among congenitally blind and sighted individuals

We examined the contribution of first-person sensory experience to concepts by comparing the meanings of perception (visual/tactile) and emission (light/sound) verbs among congenitally blind (N = 25) and sighted speakers (N = 22). Participants judged semantic similarity for pairs of verbs referring to events of visual (e.g. to peek), tactile (e.g. to feel) and amodal perception (e.g. to perceive) as well as light (e.g. to shimmer) and sound (e.g. to boom) emission and manner of motion (to roll) (total word pairs, N = 2041). Relative to the sighted, blind speakers had higher agreement among themselves on touch perception and sound emission verbs. However, for visual verbs, the judgments of blind and sighted participants were indistinguishable, both in the semantic criteria used and subject-wise variability. Blind and sighted individuals alike differentiate visual perception verbs from verbs of touch and amodal perception and differentiate among acts of visual perception e.g. intense/continuous from brief acts of looking (e.g. peek vs. stare). Light emission verbs are differentiated according to intensity (blaze vs. glow) and stability (blaze vs. flash). Thus detailed knowledge of visual word meanings is acquired without first-person sensory access.

Aesthetic experience and creativity in arts education: Ehrenzweig and the primal syncretistic perception of the child

The pertinence and worth of arts in Quebec primary schools vary considerably from one institution and school administration to another. In this paper it is argued that well-integrated arts education would bring a large array of pedagogical benefits to students, not the least of which is the preservation and the development of aesthetic perception and creativity. Actually, teachers can rely on the syncretistic and amodal perception of the young child, a learning mode that is analogous to the aesthetic experience and fostered by art practices. The paper draws notably on Ehrenzweig and Winnicott to advocate that the school, through arts education, should preserve and make use of this precious perceptive faculty, and temper the educational bias that largely favours the logical or analytical mode of the child. It advocates for an essential developmental balance between the child’s intuitive and analytical faculties, a healthy equilibrium that arts education has the potential to instil. Children and society both have much to gain from the school’s fostering of arts education and aesthetic perception: children from their balanced development and society from future creative citizens’ contributions.

Seeking the boundary of boundary extension

Boundary extension (BE) is a remarkably consistent visual memory error in which participants remember seeing a more wide-angle image of a scene than was actually viewed (Intraub & Richardson, Journal of Experimental Psychology: Learning, Memory, and Cognition 15:179-187, 1989). Multiple stimulus factors are thought to contribute to the occurrence of BE, including object recognition, conceptual knowledge of scenes, and amodal perception at the view boundaries (Intraub, Wiley Interdisciplinary Reviews: Cognitive Science 3:117-127, 2012). In the present study, we used abstract scenes instead of images of the real world, in order to remove expectations based on semantic associations with objects and the schematic context of the view. Close-angle and wide-angle scenes were created using irregular geometric shapes rated by independent observers as lacking any easily recognizable structure. The abstract objects were tested on either a random-dot or a blank background in order to assess the influence of implied continuation of the image beyond its boundaries. The random-dot background conditions had background occlusion cues either present or absent at the image border, in order to test their influence on BE in the absence of high-level information about the scenes. The results indicate that high-level information about objects and schematic context is unnecessary for BE to occur, and that occlusion information at the image boundary also has little influence on BE. Contrary to previous studies, we also found clear BE for all conditions, despite using scenes depicting undetailed objects on a blank white background. The results highlighted the ubiquitous nature of BE and the adaptability of scene perception processes.

Representing the impossible

A theory of perception must be capable of explaining the full range of conscious perception, including amodal perception. In amodal perception we perceive the world to contain physical features that are not directly detectable by the sensory receptors. According to the active-externalist theory of perception, amodal perception depends on active engagement with perceptual objects. This paper focuses on amodal visual perception and presents a counter-example to the idea that active-externalism can account for amodal perception. The counter-example involves the experience of so-called “impossible objects,” objects experienced in perceptual character as having geometrical properties that no physically real object can have.

MENTAL IMAGERY AND THE VARIETIES OF AMODAL PERCEPTION

AbstractThe problem of amodal perception is the problem of how we represent features of perceived objects that are occluded or otherwise hidden from us. Bence Nanay (2010) has recently proposed that we amodally perceive an object's occluded features by imaginatively projecting them into the relevant regions of visual egocentric space. In this paper, I argue that amodal perception is not a single, unitary capacity. Drawing appropriate distinctions reveals amodal perception to be characterized not only by mental imagery, as Nanay suggests, but also by genuinely visual representations as well as beliefs. I conclude with some brief remarks on the role of object‐directed bodily action in conferring a sense of unseen presence on an object's occluded features.

Rethinking visual scene perception.

A classic puzzle in understanding visual scene perception is how to reconcile the physiological constraints of vision with the phenomenology of seeing. Vision captures information via discrete eye fixations, interrupted by saccadic suppression, and limited by retinal inhomogeneity. Yet scenes are effortlessly perceived as coherent, continuous, and meaningful. Two conceptualizations of scene representation will be contrasted. The traditional visual-cognitive model casts visual scene representation as an imperfect reflection of the visual sensory input alone. By contrast, a new multisource model casts visual scene representation in terms of an egocentric spatial framework that is 'filled-in' by visual sensory input, but also by amodal perception, and by expectations and by constraints derived from rapid-scene classification and object-to-context associations. Together, these nonvisual sources serve to 'simulate' a likely surrounding scene that the visual input only partially reveals. Pros and cons of these alternative views will be discussed. WIREs Cogn Sci 2012, 3:117-127. doi: 10.1002/wcs.149 For further resources related to this article, please visit the WIREs website.

Spatiotemporal Boundary Formation: Similar Performance for Modal and Amodal Completion

There is a controversy about whether modal and amodal completion are driven by the same or by different mechanisms. In this sense, there are experimental and theoretical evidence in the literature supporting both hypotheses. In this study, we present the results of three experiments showing that performance for modal and amodal completion in the case of Spatiotemporal Boundary Formation (SBF) are similar. We use the type of stimuli used Shipley and Kellman (1993) to explore SBF and introduced disparity to produce modal and amodal percepts. In the first experiment we performed a Vernier acuity task between two motion-defined bars and tested the threshold for a range of disparities (+3, +1, 0, -1, -3), speeds (1, 2, 4 deg/s), and dot densities (4 and 8 dots/deg2). Results show that Vernier threshold does not depend on the amount of disparity for all experimental conditions. Moreover, there are no significant differences between thresholds obtained with positive and negative disparities, which would suggest a similarity between modality and amodallity. We wondered whether this result would keep the same for other tasks used in previous experiments such as form discrimination. In the second experiment, we used the same type of stimulus but asked the observers to discriminate between thinned and fattened bars. Results show no differences between modal and amodal perception, consistently with the first experiment. In the third experiment we measured Vernier thresholds for three temporal windows (50, 100, 350 ms), since it was found, in previous studies, some differences between modal and amodal perception for short stimulus presentation. Results show that for 100 ms four of six subjects present slight differences between thresholds obtained with positive and negative disparities. These results show that mechanisms underlying modal and amodal completion have the same efficiency, which could suggests that they are actually the same.

On amodal perception and language in art therapy with autism

This paper is based on the analysis of artwork produced by a 36-year-old male client with mild learning disability, who was withdrawn or encapsulated for self-protection into mild ‘shell-type’ autism. He was offered weekly individual psychodynamic art therapy for a two-year period in all. The analysis was carried out a year later as an MA research project, using two frameworks appropriate to non-verbal aspects of art therapy, one theoretical and the other visual. The findings offer evidence in art therapy of a language consisting of sensations and perceptions registered and expressed by the body, and between bodies, as contours, inscriptions or signs of expression. This language is structural, giving shape and form to sensed perceptions according to an innate temporal and spatial ordering. Its dynamic process constructs and develops the sense of self and other and underpins the capacity for intersubjective relatedness and learning.

Perception and imagination: amodal perception as mental imagery

When we see an object, we also represent those parts of it that are not visible. The question is how we represent them: this is the problem of amodal perception. I will consider three possible accounts: (a) we see them, (b) we have non-perceptual beliefs about them and (c) we have immediate perceptual access to them, and point out that all of these views face both empirical and conceptual objections. I suggest and defend a fourth account, according to which we represent the occluded parts of perceived objects by means of mental imagery. This conclusion could be thought of as a (weak) version of the Strawsonian dictum, according to which “imagination is a necessary ingredient of perception itself”.

Reflecting on Words and Letters from the Perspective of Embodiment, with Commentary on Essays by Daphne Lei, Susan Phillips, and Sohini Ray

Reflecting on Words and Letters from the Perspective of Embodiment, with Commentary on Essays by Daphne Lei, Susan Phillips, and Sohini Ray Deidre Sklar I. A dance ethnographer inevitably raises her hand to speak up for movement as the mediating factor in the relationship between writing and embodiment. Human movement creates inscriptions, whether with stone tools, pens and pencils, linotype machines, or computers. Orthographic systems travel through our bodies not only as symbols but as physical impulses like muscle contractions and extensions, creating complex movements and kinesthetic sensations. While European and American philosophers have long been concerned with the visual perception of movement, they have largely ignored awareness of movement sensation, or kinesthesia, as an epistemological mode. Omitted from the sensorium, [End Page 155] kinesthesia is nonetheless a primary means by which we know, engage with, and make sense of the world. My discussion of orthography and embodiment builds upon this premise to suggest ways that language implicates kinesthesia, thereby troubling the notion of a disconnect between disembodied sign and sensate body. I take as point of departure Elliot R. Wolfson's exegesis of kabbala in terms of " textual embodiment," originally presented along with several of this volume's essays at a University of California Humanities Research Institute seminar on "Gesture and Inscription" in 2002.1 This will launch my consideration of the essays by Daphne Lei, Susan Phillips and Sohini Ray in this volume. According to Wolfson, in kabbalah, "The three books by which God created the world allude to the congruence of thought, speech and writing" (2005:204). Whereas for God, words, whether thought, spoken, or written, "constitute the very essence of things they name, whence derives the creative potency of language," for humans, "words at best are 'signs' or 'symbols' that point to the things they name but not to their essence" (204). Thus Wolfson considers that in kabbala, humans are distinguished from God on the basis of the conventionality of language. Humans do not create the world in their use of language; they reiterate, by convention, what has already been created. Where do we humans get the idea that God has the capacity, in converging a sound, a mark, and a thought, to create the phenomenon to which that convergence refers, except that we, too, can sample this experience? I suggest that, in terms of human capacity, what is being brought out here is the difference between the conventional, iterative, and automatic use of language and an awareness, while speaking or writing, that we are bringing to life—at least in somato-mental consciousness—the thought represented by the sounding and marking of speech and orthography. It is not just that in internally hearing or picturing the combination of sound or letter symbols that constitute words we conjure up and imbue with vitality a somato-mental experience of a thought, though that in itself is wondrous; we also bring to life the full and changing load of associations that accumulate in the confluence of symbolic sound, mark, and thought. In short, we evoke a world. To be aware of this peira, or sampling, of the creativity inherent in language is to be aware of a miracle inherent in our natures, an "I can," to use Merleau-Ponty's term, of being human. In kabbalah, God creates the world through the confluence of sound (speech), image (writing) and thought; phenomenologically, we humans have the capacity to [End Page 156] realize, indeed, repeat, that creative act by piercing through the conventionality of language to an awareness of the multi-sensory and world-building miraculousness of the language process itself. We know from recent studies in child development that we are born with the capacity for what child psychologist Daniel Stern calls "amodal perception" (1985:51), the ability to translate information between sensory modalities. Infants can recognize visually an object they previously knew only through touch, translate sound intensities (loudness) to visual intensities (brightness) and temporal patterns (beat, rhythm, duration) between visual and auditory modes. Infants do this before they can recognize or identify objects, including themselves as objects, or "selves." The capacity for amodal recognition of qualities precedes language. Philosopher Mark Johnson (1987) argues that we form pre...

The Developmental Trajectory from Amodal Perception to Empathy and Communication: The Role of Mirror Neurons in This Process

In a recent paper entitled “Mirror Neurons. Procedural Learning and the Positive New Experience” (Wolf et al., 2000), data were presented about a special type of neuron, the mirror neuron, originally located by Rizzolatti and his colleagues (1995). These neurons were discussed as they related to a particular developmental view of psychoanalysis, developmental systems self psychology (Shane, Shane, and Gales. 1997). In this paper, we focus on how this mirror neuron system might contribute to the development of communicative abilities in humans. First we summarize the research findings about mirror neurons and how they apply to humans. We then attempt to demonstrate how the mirror neuron system might be involved in a developmental sequence hypothesized by Kohut (1984), Stern (1985), and others to begin in infancy. We postulate that this trajectory starts with the onset of “amodal perception” (Stern, 1985) and then proceeds to affect resonance, joint attention, and ultimately to symbolization of language. In this paper, we attempt to integrate these concepts with a formulation of empathy and demonstrate what might go awry in developmental disorders when the normative sequence of development described above does not take place.

Infant observation, experimental infant research and psychodynamic theory regarding lack of self/other differentiation

In this paper I draw upon infant observation, experimental infant research and psy chodynamic theory to explore the early development of internal objects in which there is no differentiation between self and object. I have used an infant observation to experiment with the idea that, early on, identificatory processes predominate over projective mechanisms. From developmentalist understanding of early perception, I have suggested that amodal perception may create a different kind of object than featural perception. With amodally shaped objects/representations the difference between self and other is irrelevant, and with feature-based objects/representations what counts is the difference.

The development of communication and language in deaf and severely hard of hearing children: implications for the future

Severe hearing impairment is seldom detected in children before the age of 6–12 months as parent–infant interaction is similar to that of a normal parent–child interaction. This is probably due to an innate capacity of infants to take information in one sensory modality and translate it into another, called amodal perception. The roots of language are traced to early proto-conversations, as well as to early pretend play. Relationships are viewed as the context in which socialisation takes place, basic competences emerge, regulations of emotions develop and communication skills are acquired. If habilitation after diagnosis of a severe hearing impairment primarily is focused on an oral–aural approach, natural patterns of communication between parent and child will gradually disappear, which will have negative implications on the development of these children. If, instead, they are allowed to develop those means of communication that are easy for them to produce and to perceive, positive consequences have been registered on the development of communication and language, as well as on their socio-emotional and cognitive development. When these children have been given opportunities to become bilingual with a signed and a written and/or spoken language, it has enabled them to attend higher education, to have a qualified job and thereby a good life in the future.

Perceptions Seeking Consciousness: A Developmental Model for Interpreting the Origins of Reality

Abstract A repertoire of developmental capacities facilitates the infant's understanding of reality, including: categorization, amodal perception, effectance motivation, attachment behavior, primary and secondary processes. A newly described developmental phenomenon—previewing is the strategy by which the caregiver introduces the infant to upcoming developmental skill and the interpersonal implications this skill will have for the dyadic relationship. Through previewing, the infant comes to understand that seemingly uncontrollable developmental changes as well as changes in subjective perception can be mastered. Because it is derived from early life experience, previewing may also contribute to our understanding of perceptions of adults. This paper discusses the developmental proclivities that characterize the perception of reality.