Spotting animals in natural scenes: efficiency of humans and monkeys at very low contrasts

The ability of monkeys to categorize objects in visual stimuli such as natural scenes might rely on sets of low-level visual cues without any underlying conceptual abilities. Using a go/no-go rapid animal/non-animal categorization task with briefly flashed achromatic natural scenes, we show that both human and monkey performance is very robust to large variations of stimulus luminance and contrast. When mean luminance was increased or decreased by 25–50%, accuracy and speed impairments were small. The largest impairment was found at the highest luminance value with monkeys being mainly impaired in accuracy (drop of 6% correct vs. <1.5% in humans), whereas humans were mainly impaired in reaction time (20 ms increase in median reaction time vs. 4 ms in monkeys). Contrast reductions induced a large deterioration of image definition, but performance was again remarkably robust. Subjects scored well above chance level, even when the contrast was only 12% of the original photographs (≈81% correct in monkeys; ≈79% correct in humans). Accuracy decreased with contrast reduction but only reached chance level -in both species- for the most extreme condition, when only 3% of the original contrast remained. A progressive reaction time increase was observed that reached 72 ms in monkeys and 66 ms in humans. These results demonstrate the remarkable robustness of the primate visual system in processing objects in natural scenes with large random variations in luminance and contrast. They illustrate the similarity with which performance is impaired in monkeys and humans with such stimulus manipulations. They finally show that in an animal categorization task, the performance of both monkeys and humans is largely independent of cues relying on global luminance or the fine definition of stimuli.     

Sellars and Nonconceptual Content

In this paper I take up the question of whether Wilfrid Sellars has a notion of non‐conceptual perceptual content. The question is controversial, being one of the fault lines along which so‐called left and right Sellarsians diverge. In the paper I try to make clear what it is in Sellars' thought that leads interpreters to such disparate conclusions. My account depends on highlighting the importance of Sellars' little discussed thesis that perception involves a systematic form of mis‐categorization, one where perceivers mistake their sensory states to be properties of physical objects. I argue that the counterpart color and shape attributes of these states, which become ‘point of viewish’ when organized by the productive imagination, provides perceptual experience with its non‐conceptual representational content. I then argue that this content is not a form of the mythical Given because one can only have a non‐conceptual point of view on an object when an object is introduced into one's perceptual experience through the conceptual mis‐taking of one's sensory states. So, while Sellars has a notion of non‐conceptual representational content, it can only be salient in the context of a perceptual act that is conceptual overall.     

Categorization difficulty is associated with negative evaluation in the “uncanny valley” phenomenon

Human observers often experience strongly negative impressions of human‐like objects falling within a particular range of visual similarity to real humans (the “uncanny valley” phenomenon). We hypothesized that negative impressions in the uncanny valley phenomenon are related to a difficulty in object categorization. We produced stimulus images by morphing two of each of real, stuffed, and cartoon human face images (Experiment 1). Observers were asked to categorize each of these images as either category and evaluated the likability of the image. The results revealed that the longest latency, the highest ambiguity in categorization, and the lowest likability score co‐occurred at consistent morphing percentages. Similar results were obtained even when we employed stimulus images that were created by morphing two of each of real, stuffed, and cartoon dog images (Experiment 2). However, the effect of categorization difficulty on evaluation was weak when two real human faces were morphed (Experiment 3). These results suggest that the difficulty in categorizing an object as either of two dissimilar categories is linked to negative evaluation regardless of whether the object is human‐related or not.     

The lasting effects of process‐specific versus stimulus‐specific learning during infancy

The capacity to tell the difference between two faces within an infrequently experienced face group (e.g. other species, other race) declines from 6 to 9 months of age unless infants learn to match these faces with individual‐level names. Similarly, the use of individual‐level labels can also facilitate differentiation of a group of non‐face objects (strollers). This early learning leads to increased neural specialization for previously unfamiliar face or object groups. The current investigation aimed to determine whether early conceptual learning between 6 and 9 months leads to sustained behavioral advantages and neural changes in these same children at 4–6 years of age. Results suggest that relative to a control group of children with no previous training and to children with infant category‐level naming experience, children with early individual‐level training exhibited faster response times to human faces. Further, individual‐level training with a face group – but not an object group – led to more adult‐like neural responses for human faces. These results suggest that early individual‐level learning results in long‐lasting process‐specific effects, which benefit categories that continue to be perceived and recognized at the individual level (e.g. human faces).     

Infants' attribution of a goal to a morphologically unfamiliar agent

How do infants identify the psychological actors in their environments? Three groups of 12‐month‐old infants were tested for their willingness to encode a simple approach behavior as goal‐directed as a function of whether it was performed by (1) a human hand, (2) a morphologically unfamiliar green object that interacted with a confederate and behaved intentionally, or (3) the same unfamiliar green object that behaved in a matched, but apparently random manner. Using a visual habituation technique, only infants in the first two conditions were found to encode the approach behavior as goal‐directed. Thus infants appear able to attribute goals to non‐human, even unfamiliar agents. These results imply that by the end of the first year of life infants have a broad notion of what counts as an agent that cannot easily be reduced to humans, objects that are perceptually similar to humans, or objects that display self‐propulsion.     

Self‐generated variability in object images predicts vocabulary growth

Object names are a major component of early vocabularies and learning object names depends on being able to visually recognize objects in the world. However, the fundamental visual challenge of the moment‐to‐moment variations in object appearances that learners must resolve has received little attention in word learning research. Here we provide the first evidence that image‐level object variability matters and may be the link that connects infant object manipulation to vocabulary development. Using head‐mounted eye tracking, the present study objectively measured individual differences in the moment‐to‐moment variability of visual instances of the same object, from infants’ first‐person views. Infants who generated more variable visual object images through manual object manipulation at 15 months of age experienced greater vocabulary growth over the next six months. Elucidating infants’ everyday visual experiences with objects may constitute a crucial missing link in our understanding of the developmental trajectory of object name learning.     

Do chimpanzees anticipate an object’s weight? A field experiment on the kinematics of hammer-lifting movements in the nut-cracking Taï chimpanzees

When humans are about to manipulate an object, our brains use visual cues to recall an internal representation to predict its weight and scale the lifting force accordingly. Such a long-term force profile, formed through repeated experiences with similar objects, has been proposed to improve manipulative performance. Skillful object manipulation is crucial for many animals, particularly those that rely on tools for foraging. However, despite enduring interest in tool use in non-human animals, there has been very little investigation of their ability to form an expectation about an object’s weight. In this study, we tested whether wild chimpanzees use long-term force profiles to anticipate the weight of a nut-cracking hammer from its size. To this end, we conducted a field experiment presenting chimpanzees with natural wooden hammers and artificially hollowed, lighter hammers of the same size and external appearance. We used calibrated videos from camera traps to extract kinematic parameters of lifting movements. We found that, when lacking previous experience, chimpanzees lifted hollowed hammers with a higher acceleration than natural hammers (overshoot effect). After using a hammer to crack open one nut, chimpanzees tuned down the lifting acceleration for the hollowed hammers, but continued lifting natural hammers with the same acceleration. Our results show that chimpanzees anticipate the weight of an object using long-term force profiles and suggest that, similarly to humans, they use internal representations of weight to plan their lifting movements.     

Learning to perceive object unity: a connectionist account

To explore questions of how human infants begin to perceive partly occluded objects, we devised two connectionist models of perceptual development. The models were endowed with an existing ability to detect several kinds of visual information that have been found important in infants’ and adults’ perception of object unity (motion, co‐motion, common motion, relatability, parallelism, texture and T‐junctions). They were then presented with stimuli consisting of either one or two objects and an occluding screen. The models’ task was to determine whether the object or objects were joined when such a percept was ambiguous, after specified amounts of training with events in which a subset of possible visual information was provided. The model that was trained in an enriched environment achieved superior levels of performance and was able to generalize veridical percepts to a wide range of novel stimuli. Implications for perceptual development in humans, current theories of development and origins of knowledge are discussed.     

Figural goodness,stimulus heterogeneity,similarity and object segregation in infancy

The segregation of objects from other objects in visual arrays is a fundamental function of our visual system. Research suggests that adults’ detection of a target among nontargets is affected by the heterogeneity of array elements and the resulting changes in target–nontarget and nontarget–nontarget similarities. We examined the effects of heterogeneity and similarity on object segregation in infancy. In Experiment 1, 5.5‐month‐olds detected a misoriented element in an array when the array elements were spatially arranged in a ‘good’ configuration but not when they were arranged in a ‘poor’ configuration. In Experiment 2, infants detected a vertical line in a homogeneous array of 55° or 125° lines, but failed to do so in a heterogeneous array of 55° and 125° lines. Thus, heterogeneity in both the arrangement and identity of array elements affected infants’ discrepancy detection. Because the average target–nontarget similarity was the same in the two conditions of Experiment 2, the results also indicated that nontarget–nontarget similarity independently affects discrepancy detection in infancy. These results are consistent with models of object segregation by adults, and suggest that stimulus heterogeneity and similarity have analogous effects on object segregation at 5.5 months of age and in adulthood.     

THIS ARTICLE HAS BEEN RETRACTED: Enumeration of objects and substances in non‐human primates: experiments with brown lemurs (Eulemur fulvus)

Both human infants and adult non‐human primates share the capacity to track small numbers of objects across time and occlusion. The question now facing developmental and comparative psychologists is whether similar mechanisms give rise to this capacity across the two populations. Here, we explore whether non‐human primates’ object tracking abilities are subject to the same constraints as those of human infants. In particular, we examine whether one primate species, the brown lemur (Eulemur fulvus), also fails to represent and enumerate objects when they behave non‐rigidly or non‐cohesively. We presented lemurs with a series of expectancy violation studies involving simple 1 + 1 addition events in which we varied the entities to be enumerated. Like infants, lemurs successfully enumerated the two objects when those objects were rigid, cohesive individuals, but failed to enumerate similar‐looking non‐rigid piles of sand. In contrast to human infants, however, lemurs successfully enumerated non‐cohesive objects that broke into multiple pieces. These results are discussed in light of recent theories about object processing in human infants and adults.     

Functional and dynamic properties of visual peripersonal space

In order to code visual peripersonal space, human and non-human primates need an integrated system that controls both visual and tactile inputs within peripersonal space around the face and the hand, based on visual experience of body parts. The existence of such a system in humans has been demonstrated, and there is evidence showing that visual peripersonal space relating to the hand has important dynamic properties, for example, it can be expanded and contracted depending on tool use. There is also evidence for a high degree of functional similarity between the characteristics of the visual peripersonal space in humans and in monkeys.     

How do 8‐month‐old infants recognize causality in object motion and that in human action?1

Abstract: In Experiment 1, 8‐month‐old infants were first habituated to the event in which a moving object collided with another behind an occluder, then they were shown the two test events with no occluder: the contact event, in which the two objects actually collided, and the non‐contact event, in which the second object started to move without contact with the first. The infants looked at both events for an equal amount of time. In Experiment 2, in which the first object was a human actor, however, infants looked at the non‐contact event reliably longer than the contact event. In Experiment 3, in which both objects were human actors stood face‐to‐back, infants looked at the non‐contact event longer, whereas in Experiment 4, in which human actors faced toward each other, infants looked at both events equally. In Experiment 5, in which the first actor told the second to go, 10‐month‐old infants looked at both events for an equal amount of time. These results suggest that 8‐ and 10‐month‐old infants appreciate different causal principles between objects and humans, and that, in doing this, they may acknowledge the possibility of communication between humans.     

Evidence for complete translational and reflectional invariance in visual object priming.

The magnitude of priming on naming reaction times and on the error rates, resulting from the perception of a briefly presented picture of an object approximately 7 min before the primed object, was found to be independent of whether the primed object was originally viewed in the same hemifield, left-right or upper-lower, or in the same left-right orientation. Performance for same-name, different-examplar images was worse than for identical images, indicating that not only was there priming from block one to block two, but that some of the priming was visual, rather than purely verbal or conceptual. These results provide evidence for complete translational and reflectional invariance in the representation of objects for purposes of visual recognition. Explicit recognition memory for position and orientation was above chance, suggesting that the representation of objects for recognition is independent of the representations of the location and left-right orientation of objects in space.     

Viewpoint-dependent recognition of familiar faces

The question whether object representations in the human brain are object-centered or viewer-centered has motivated a variety of experiments with divergent results. A key issue concerns the visual recognition of objects seen from novel views. If recognition performance depends on whether a particular view has been seen before, it can be interpreted as evidence for a viewer-centered representation. Earlier experiments used unfamiliar objects to provide the experimenter with complete control over the observer's previous experience with the object. In this study, we tested whether human recognition shows viewpoint dependence for the highly familiar faces of well-known colleagues and for the observer's own face. We found that observers are poorer at recognizing their own profile, whereas there is no difference in response time between frontal and profile views of other faces. This result shows that extensive experience and familiarity with one's own face is not sufficient to produce viewpoint invariance. Our result provides strong evidence for viewer-centered representations in human visual recognition even for highly familiar objects.     

A test of the embodied simulation theory of object perception: potentiation of responses to artifacts and animals

Theories of embodied object representation predict a tight association between sensorimotor processes and visual processing of manipulable objects. Previous research has shown that object handles can ‘potentiate’ a manual response (i.e., button press) to a congruent location. This potentiation effect is taken as evidence that objects automatically evoke sensorimotor simulations in response to the visual presentation of manipulable objects. In the present series of experiments, we investigated a critical prediction of the theory of embodied object representations that potentiation effects should be observed with manipulable artifacts but not non-manipulable animals. In four experiments we show that (a) potentiation effects are observed with animals and artifacts; (b) potentiation effects depend on the absolute size of the objects and (c) task context influences the presence/absence of potentiation effects. We conclude that potentiation effects do not provide evidence for embodied object representations, but are suggestive of a more general stimulus–response compatibility effect that may depend on the distribution of attention to different object features.     

Similarity of the perimeters in the Ebbinghaus illusion

Coren and Miller (1974) and Coren and Enns (1993) argued that the magnitude of the Ebbinghaus illusion is a function of the rated or conceptual similarity of the inducing objects to the test object. In three experiments, we examined the convergence between conceptual similarity and illusion magnitude. The first failed to find support for this parallel. Two further experiments yielded support for an alternative hypothesis that the magnitude of the Ebbinghaus illusion is a function of the similarity of the perimeters of the inducing object to the test object. The similarity of the centers had no effect. These results suggest that the information used to estimate size is computed earlier in the visual system than suggested by Coren and colleagues and apparently does not involve the use of conceptual information.     

Dissociation of visual localization and visual detection in rhesus monkeys (Macaca mulatta)

Conscious and unconscious cognitive processes contribute independently to human behavior and can be dissociated. For example, humans report failing to see objects clearly in the periphery while simultaneously being able to grasp those objects accurately (Milner in Proc R Soc B Biol Sci 279:2289–2298, 2012). Knowing whether similar dissociations are present in nonverbal species is critical to our understanding of comparative psychology and the evolution of brains. However, such dissociations are difficult to detect in nonhumans because verbal reports of experience are the main way we discriminate putative conscious from unconscious processing. We trained monkeys in a localization task in which they responded to the location where a target appeared, and a matched detection task in which they reported the presence or absence of the same target. We used masking to manipulate the visibility of targets. Accuracy was high in both tasks when stimuli were unmasked and was attenuated by visual masking. At the strongest level of masking, performance in the detection task was at chance, while localization remained significantly above chance. Critically, errors in the detection task were predominantly misses, indicating that the monkeys’ behavior remained under stimulus control, but that the monkeys did not detect the target despite above-chance localization. While these results cannot establish the existence of phenomenal vision in monkeys, the dissociation of visually guided action from detection parallels the dissociation of conscious and unconscious vision seen in humans.     

Using automated controlled rearing to explore the origins of object permanence

What are the origins of object permanence? Despite widespread interest in this question, methodological barriers have prevented detailed analysis of how experience shapes the development of object permanence in newborn organisms. Here, we introduce an automated controlled‐rearing method for studying the emergence of object permanence in strictly controlled virtual environments. We used newborn chicks as an animal model and recorded their behavior continuously (24/7) from the onset of vision. Across four experiments, we found that object permanence can develop rapidly, within the first few days of life. This ability developed even when chicks were reared in impoverished visual environments containing no object occlusion events. Object permanence failed to develop, however, when chicks were reared in environments containing temporally non‐smooth objects (objects moving on discontinuous spatiotemporal paths). These results suggest that experience with temporally smooth objects facilitates the development of object permanence, confirming a key prediction of temporal learning models in computational neuroscience.     

类别表征与策略的脑科学研究及对思维教学的启示

脑科学研究表明,分类中对物体的知觉表征激活颞下叶皮层和颞中区等区域;语义表征激活前额皮层等。语义加工和知觉表征相互作用的脑机制表现为额区与视觉区域间信息的传递、提取。分类策略主要有规则策略和相似性策略两种。规则策略激活额区等广泛区域;相似性策略与视觉区域相关,包含着对个别样例的记忆过程。基于以上研究结果,本文从学习材料的表征和对材料进行表征时的策略两个方面为思维教学提出了一些建议。     

Object memory and change detection: dissociation as a function of visual and conceptual similarity

People often fail to detect a change between two visual scenes, a phenomenon referred to as change blindness. This study investigates how a post-change object's similarity to the pre-change object influences memory of the pre-change object and affects change detection. The results of Experiment 1 showed that similarity lowered detection sensitivity but did not affect the speed of identifying the pre-change object, suggesting that similarity between the pre- and post-change objects does not degrade the pre-change representation. Identification speed for the pre-change object was faster than naming the new object regardless of detection accuracy. Similarity also decreased detection sensitivity in Experiment 2 but improved the recognition of the pre-change object under both correct detection and detection failure. The similarity effect on recognition was greatly reduced when 20% of each pre-change stimulus was masked by random dots in Experiment 3. Together the results suggest that the level of pre-change representation under detection failure is equivalent to the level under correct detection and that the pre-change representation is almost complete. Similarity lowers detection sensitivity but improves explicit access in recognition. Dissociation arises between recognition and change detection as the two judgments rely on the match-to-mismatch signal and mismatch-to-match signal, respectively.