Dataset Information

Invariant texture perception is harder with synthetic textures: Implications for models of texture processing.

ABSTRACT: Texture synthesis models have become a popular tool for studying the representations supporting texture processing in human vision. In particular, the summary statistics implemented in the Portilla-Simoncelli (P-S) model support high-quality synthesis of natural textures, account for performance in crowding and search tasks, and may account for the response properties of V2 neurons. We chose to investigate whether or not these summary statistics are also sufficient to support texture discrimination in a task that required illumination invariance. Our observers performed a match-to-sample task using natural textures photographed with either diffuse overhead lighting or lighting from the side. Following a briefly presented sample texture, participants identified which of two test images depicted the same texture. In the illumination change condition, illumination differed between the sample and the matching test image. In the no change condition, sample textures and matching test images were identical. Critically, we generated synthetic versions of these images using the P-S model and also tested participants with these. If the statistics in the P-S model are sufficient for invariant texture perception, performance with synthetic images should not differ from performance in the original task. Instead, we found a significant cost of applying texture synthesis in both lighting conditions. We also observed this effect when power-spectra were matched across images (Experiment 2) and when sample and test images were drawn from unique locations in the parent textures to minimize the contribution of image-based processing (Experiment 3). Invariant texture processing thus depends upon measurements not implemented in the P-S algorithm.

SUBMITTER: Balas B

PROVIDER: S-EPMC4529380 | biostudies-literature | 2015 Oct

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Invariant texture perception is harder with synthetic textures: Implications for models of texture processing.

Balas Benjamin B Conlin Catherine C

Vision research 20150207 Pt B

Texture synthesis models have become a popular tool for studying the representations supporting texture processing in human vision. In particular, the summary statistics implemented in the Portilla-Simoncelli (P-S) model support high-quality synthesis of natural textures, account for performance in crowding and search tasks, and may account for the response properties of V2 neurons. We chose to investigate whether or not these summary statistics are also sufficient to support texture discriminat ...[more]

PMID: 25668773

Similar Datasets

Project description:Although the natural haptic perception of textures includes active finger movements, it is unclear how closely perception and movements are linked. Here we investigated this question using oriented textures. Textures that are composed of periodically repeating grooves have a clear orientation defined by the grooves. The direction of finger movement relative to texture orientation determines the availability of temporal cues to the spatial period of the texture. These cues are absent during movements directed in line with texture orientation, whereas movements orthogonal to texture orientation maximize the temporal frequency of stimulation. This may optimize temporal cues. In Experiment 1 we tested whether texture perception gets more precise the more orthogonal the movement direction is to the texture. We systematically varied the movement direction within a 2IFC spatial period discrimination task. As expected, perception was more precise (lower discrimination thresholds) when finger movements were directed closer towards the texture orthogonal as compared to in parallel to the texture. In Experiment 2 we investigated whether people adjust movement directions to the texture orthogonal in free exploration. We recorded movement directions during free exploration of standard and comparison gratings. The standard gratings were clearly oriented. The comparison gratings did not have a clear orientation defined by grooves. Participants adjusted movement directions to the texture orthogonal only for clearly oriented textures (standards). The adjustment to texture orthogonal was present in the final movement but not in the first movement. This suggests that movement adjustment is based on sensory signals for texture orientation that were gathered over the course of exploration. In Experiment 3 we assessed whether the perception of texture orientation and movement adjustments are based on shared sensory signals. We determined perceptual thresholds for orientation discrimination and computed 'movometric' thresholds from the stroke-by-stroke adjustment of movement direction. Perception and movements were influenced by a common factor, the spatial period, suggesting that the same sensory signals for texture orientation contribute to both. We conclude that people optimize texture perception by adjusting their movements in directions that maximize temporal cue frequency. Adjustments are performed on the basis of sensory signals that are also used for perception.

Dataset Information

Invariant texture perception is harder with synthetic textures: Implications for models of texture processing.

Publications

Invariant texture perception is harder with synthetic textures: Implications for models of texture processing.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure