Limits...
Diagnostic spatial frequencies and human efficiency for discriminating actions.

Thurman SM, Grossman ED - Atten Percept Psychophys (2011)

Bottom Line: The first, more traditional, approach measured signal-to-noise ratio (s/n) thresholds for videos filtered by one of six Gaussian band-pass filters ranging from 4 to 128 cycles/image.The second approach used SF "bubbles", Willenbockel et al. (Journal of Experimental Psychology.Efficiency on this task was estimated by comparing s/n thresholds for humans to an ideal observer, and was estimated to be quite low (>.04%) for both experiments.

View Article: PubMed Central - PubMed

Affiliation: Department of Cognitive Sciences, University of California, Irvine, 3151 Social Science Plaza, Irvine, CA 92697-5100, USA. sthurman@uci.edu

ABSTRACT
Humans extract visual information from the world through spatial frequency (SF) channels that are sensitive to different scales of light-dark fluctuations across visual space. Using two methods, we measured human SF tuning for discriminating videos of human actions (walking, running, skipping and jumping). The first, more traditional, approach measured signal-to-noise ratio (s/n) thresholds for videos filtered by one of six Gaussian band-pass filters ranging from 4 to 128 cycles/image. The second approach used SF "bubbles", Willenbockel et al. (Journal of Experimental Psychology. Human Perception and Performance, 36(1), 122-135, 2010), which randomly filters the entire SF domain on each trial and uses reverse correlation to estimate SF tuning. Results from both methods were consistent and revealed a diagnostic SF band centered between 12-16 cycles/image (about 1-1.25 cycles/body width). Efficiency on this task was estimated by comparing s/n thresholds for humans to an ideal observer, and was estimated to be quite low (>.04%) for both experiments.

Show MeSH

Related in: MedlinePlus

SF filters, example stimuli and example noise fields from Experiments 1 and 2. a. Transfer functions for each of the Gaussian band-pass filters used in Experiment 1, labeled as bp1 through bp 6 from low to high spatial frequencies. b. Selected frames from an example action video (left), filtered with the darkened filter in plot A (bp 2), plus an example of dynamic Gaussian noise filtered in the same SF band (right). c. Transfer function of an example SF sampling vector using the SF bubbles method. d. Selected frames from a stimulus filtered with the transfer function in plot C (left), plus an example of dynamic Gaussian pink noise (right)
© Copyright Policy
Related In: Results  -  Collection


getmorefigures.php?uid=PMC3037475&req=5

Fig1: SF filters, example stimuli and example noise fields from Experiments 1 and 2. a. Transfer functions for each of the Gaussian band-pass filters used in Experiment 1, labeled as bp1 through bp 6 from low to high spatial frequencies. b. Selected frames from an example action video (left), filtered with the darkened filter in plot A (bp 2), plus an example of dynamic Gaussian noise filtered in the same SF band (right). c. Transfer function of an example SF sampling vector using the SF bubbles method. d. Selected frames from a stimulus filtered with the transfer function in plot C (left), plus an example of dynamic Gaussian pink noise (right)

Mentions: The action stimuli were filtered with one of six Gaussian band-pass filters, each separated by one octave with a standard deviation of 0.5 octaves. The transfer functions of the filters are displayed in Fig. 1a. The centers of the filters were 4, 8, 16, 32, 64, and 128 (high-pass) cycles/image, which corresponded to center frequencies of 0.28, 0.57, 1.13, 2.27, 4.54, 9.08 cycles/degree visual angle. We created Gaussian noise fields by drawing independent samples from a Gaussian distribution (mean = 0, SD = 1) for each pixel in a 256 x 256 array. The noise fields were then filtered with one of the six band-pass filters, creating six sets of filtered Gaussian noise. Each set contained 100 unique filtered noise fields, and dynamic noise was created by randomly choosing 25 frames from the set of 100 on each trial.Fig. 1


Diagnostic spatial frequencies and human efficiency for discriminating actions.

Thurman SM, Grossman ED - Atten Percept Psychophys (2011)

SF filters, example stimuli and example noise fields from Experiments 1 and 2. a. Transfer functions for each of the Gaussian band-pass filters used in Experiment 1, labeled as bp1 through bp 6 from low to high spatial frequencies. b. Selected frames from an example action video (left), filtered with the darkened filter in plot A (bp 2), plus an example of dynamic Gaussian noise filtered in the same SF band (right). c. Transfer function of an example SF sampling vector using the SF bubbles method. d. Selected frames from a stimulus filtered with the transfer function in plot C (left), plus an example of dynamic Gaussian pink noise (right)
© Copyright Policy
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC3037475&req=5

Fig1: SF filters, example stimuli and example noise fields from Experiments 1 and 2. a. Transfer functions for each of the Gaussian band-pass filters used in Experiment 1, labeled as bp1 through bp 6 from low to high spatial frequencies. b. Selected frames from an example action video (left), filtered with the darkened filter in plot A (bp 2), plus an example of dynamic Gaussian noise filtered in the same SF band (right). c. Transfer function of an example SF sampling vector using the SF bubbles method. d. Selected frames from a stimulus filtered with the transfer function in plot C (left), plus an example of dynamic Gaussian pink noise (right)
Mentions: The action stimuli were filtered with one of six Gaussian band-pass filters, each separated by one octave with a standard deviation of 0.5 octaves. The transfer functions of the filters are displayed in Fig. 1a. The centers of the filters were 4, 8, 16, 32, 64, and 128 (high-pass) cycles/image, which corresponded to center frequencies of 0.28, 0.57, 1.13, 2.27, 4.54, 9.08 cycles/degree visual angle. We created Gaussian noise fields by drawing independent samples from a Gaussian distribution (mean = 0, SD = 1) for each pixel in a 256 x 256 array. The noise fields were then filtered with one of the six band-pass filters, creating six sets of filtered Gaussian noise. Each set contained 100 unique filtered noise fields, and dynamic noise was created by randomly choosing 25 frames from the set of 100 on each trial.Fig. 1

Bottom Line: The first, more traditional, approach measured signal-to-noise ratio (s/n) thresholds for videos filtered by one of six Gaussian band-pass filters ranging from 4 to 128 cycles/image.The second approach used SF "bubbles", Willenbockel et al. (Journal of Experimental Psychology.Efficiency on this task was estimated by comparing s/n thresholds for humans to an ideal observer, and was estimated to be quite low (>.04%) for both experiments.

View Article: PubMed Central - PubMed

Affiliation: Department of Cognitive Sciences, University of California, Irvine, 3151 Social Science Plaza, Irvine, CA 92697-5100, USA. sthurman@uci.edu

ABSTRACT
Humans extract visual information from the world through spatial frequency (SF) channels that are sensitive to different scales of light-dark fluctuations across visual space. Using two methods, we measured human SF tuning for discriminating videos of human actions (walking, running, skipping and jumping). The first, more traditional, approach measured signal-to-noise ratio (s/n) thresholds for videos filtered by one of six Gaussian band-pass filters ranging from 4 to 128 cycles/image. The second approach used SF "bubbles", Willenbockel et al. (Journal of Experimental Psychology. Human Perception and Performance, 36(1), 122-135, 2010), which randomly filters the entire SF domain on each trial and uses reverse correlation to estimate SF tuning. Results from both methods were consistent and revealed a diagnostic SF band centered between 12-16 cycles/image (about 1-1.25 cycles/body width). Efficiency on this task was estimated by comparing s/n thresholds for humans to an ideal observer, and was estimated to be quite low (>.04%) for both experiments.

Show MeSH
Related in: MedlinePlus