Explore chapters and articles related to this topic
Visual Object Agnosia
Published in Alexander R. Toftness, Incredible Consequences of Brain Injury, 2023
Humans are excellent at object recognition. We can recognize an endless number of pieces of furniture even though they are mostly just variations of rectangle combinations. We can identify black and white line drawings of objects even though, if you think about it, a drawing of a bucket looks very little like an actual bucket. Shapes give way to meaning in our brains almost automatically, letting us organize the endless squares, triangles, and ellipses of various sizes and orientations into a coherent universe of objects. This is a skill at which computers are noticeably inferior, being unable to reliably tell the difference between a pig, dog, or loaf of bread. So, how is it that our brains can accomplish this feat in a split second while advanced mechanical minds repeatedly fail? Not to disappoint, but researchers still do not completely agree on exactly how the human brain accomplishes this marvelous feat. What we do know is that there are a lot of ways that this process can break down.
Cortical Visual Loss
Published in Vivek Lal, A Clinical Approach to Neuro-Ophthalmic Disorders, 2023
This refers to the classic failure to match simple shapes, which implies a defect in perceiving elementary properties like curvature, surface and volume (129, 131). Examples in the literature include Mr S (132) and DF (133). Shape misperception falls along a continuum, with some patients like SMK perceiving simple shapes better than complex ones (134). The residual object recognition of patients with form agnosia is fragmentary and relies on inferences from texture and color.
Teager-Kaiser Boost Clustered Segmentation of Retinal Fundus Images for Glaucoma Detection
Published in K. Gayathri Devi, Kishore Balasubramanian, Le Anh Ngoc, Machine Learning and Deep Learning Techniques for Medical Science, 2022
P M Siva Raja, R P Sumithra, K Ramanan
Image segmentation is employed to split the image into different parts based on similar pixel characteristics. In general, image segmentation is applied in different applications such as image compression, disease identification object recognition, and so on. The technique processes an entire image is inefficient to process the whole image. Therefore, for partitioning the regions, image segmentation is utilized.
Compound truncated Poisson gamma distribution for understanding multimodal SAR intensities
Published in Journal of Applied Statistics, 2023
A. D. C. Nascimento, Leandro C. Rêgo, Jonas W. A. Silva
Several phenomena produce multimodal data in practice Cobb [5]; for instance, in image processing El-Zaart and Ziou [11] and in voice recognition Povey et al. [31]. Mixture models have been employed for describing these data McLachlan and Peel [23]. Bouguila and Ziou [2] proposed a new finite mixture model based on the generalized Dirichlet distribution. They also applied their proposal to image object recognition. Yang and Liu [37] have presented a new mixture density model based on the concept of group. Mignotte et al. [24] have proposed an unsupervised segmentation based on a hierarchical Markov random field model and Yang and Krishnan [36] provided an image segmentation method based on mixture and spatial information. However, such models impose difficult to inference and optimization (crucial steps in their adjustment) procedures because the presence of a high number of parameters. To solve this issue, we propose a three-parameter probability distribution which may describe data having multiple modes.
Immediate stress alters social and object interaction and recognition memory in nearly isogenic rat strains with differing stress reactivity
Published in Stress, 2021
Alice K. Schaack, Madaline Mocchi, Katherine J. Przybyl, Eva E. Redei
Similar to object recognition, the raw spatial investigation time was also analyzed by strain and novel placement. There was no significant main effect of strain or stress in males, but there was a placement of objects × strain interaction that significantly affected investigation time (F[1, 29] = 4.48, p < 0.05; Supplemental Figure 3). WMI stressed males did not recognize the object with the original place, as indicated by spending more time (individual p-value <0.05) investigating the object at the original place. In contrast to males, WLI control and WMI stressed females recognized the object in the original space, while both stressed WLI females and WMI control females spent the same amount of time investigating the objects at their original and novel place (Supplemental Figure 3). Statistical analysis showed that in addition to the significant main effects of placement of object, there was a significant interaction between placement, strain and stress (placement of object, F[1, 25] = 7.78, p < 0.01; placement × strain × stress, F[1, 25] = 10.30, p < 0.01; Supplemental Figure 3).
Towards the next generation of LMRA instruments: the influence of generic and specific questions during risk assessment
Published in International Journal of Occupational Safety and Ergonomics, 2021
Wouter M.P. Steijn, Dolf van der Beek, Jop Groeneweg, Anne Jansen, Wieke A. Oldenhof, Ingrid Raben
Another challenge for real-life implementation lies in how situation specific information concerning relevant risk factors can be provided to employees at the right time. This challenge might be overcome with the help of developments in object recognition algorithms, e.g., Gao et al., Redmon et al. and Wang et al. for examples of applications of object recognition [20–22], and digital technology such as digital glasses [23]. Assistance in the workplace through digital technology gives control over the amount and availability of information depending on the specific situation (one can have different work activities with a wide variety of risk factors on one day). This will support on-site decision-making by providing context-specific information. This could also be beneficial in training situations.