Method and device for improving segmentation performance to be used for detecting events by using edge loss
A technology of edge loss and edge part, applied in neural learning methods, image analysis, computer components, etc.
Active Publication Date: 2020-08-04
STRADVISION
View PDF4 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
However, it is disadvantageous in that the learning for detecting the edge portion cannot be efficiently performed, and a large amount of energy is required to restore the edge portion
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0054] The following detailed description of the invention refers to the accompanying drawings, which are shown as illustrations of specific embodiments in which the invention can be practiced. These examples are described in detail to enable those skilled in the art to practice the invention. It should be understood that the various embodiments of the invention, although different from each other, are not necessarily mutually exclusive. For example, specific shapes, structures and characteristics described herein are related to one embodiment, but can be implemented in other embodiments without departing from the spirit and scope of the invention. In addition, it should be understood that the position or arrangement of each constituent element in each disclosed embodiment may be classified without departing from the spirit and scope of the present invention. Accordingly, the following detailed description should not be read as limiting, and the scope of the invention, if pro...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention relates to a method and device for improving segmentation performance to be used for detecting events by using edge loss. A learning method for improving a segmentation performance to beused for detecting events including a pedestrian event, a vehicle event, a falling event, and a fallen event using a learning device is provided. The method includes steps of: the learning device (a)instructing k convolutional layers to generate k encoded feature maps; (b) instructing k-1 deconvolutional layers to sequentially generate k-1 decoded feature maps, wherein the learning device instructs h mask layers to refer to h original decoded feature maps outputted from h deconvolutional layers corresponding thereto and h edge feature maps generated by extracting edge parts from the h original decoded feature maps; and (c) instructing h edge loss layers to generate h edge losses by referring to the edge parts and their corresponding GTs. Further, the method allows a degree of detecting traffic sign, landmark, road marker, and the like to be increased.
Description
technical field [0001] The present invention relates to a learning method for improving segmentation performance for detection of pedestrian events, car events, drop events (events in which objects fall from a moving vehicle), drop events (objects dropped to the road), and more specifically, to a learning method for improving segmentation performance using a learning device, and a learning method and a testing method and a testing device using the same, wherein the learning device includes : (i) the first to the kth convolutional layer perform more than one convolution operation on at least one feature map corresponding to at least one training image, and output the first to the kth coded feature map respectively; (ii) the (kth) -1) To the first deconvolution layer, perform at least one deconvolution operation on the kth encoded feature map, and output the (k-1)th to the first decoded feature map respectively; (iii) the first to the hth Mask layers, respectively corresponding...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/34G06K9/46G06K9/62G06N3/04G06N3/08G06V10/764
CPCG06N3/084G06V10/26G06V10/454G06N3/045G06F18/214G06T7/12G06T2207/20081G06T2207/30252G06T2207/20084G06V10/82G06V10/764G06T7/13G06T7/11G06V10/776G06V20/56G06N3/0464G06N20/00G06N3/08G06F18/217
Inventor 金桂贤金镕重金寅洙金鹤京南云铉夫硕焄成明哲吕东勋柳宇宙张泰雄郑景中诸泓模赵浩辰
Owner STRADVISION