Methods and systems are disclosed for analyzing one or more images of a textile to determine a presence or absence of defects. In one example, an image of at least a portion of a textile may be obtained and compared to a reference image of a reference textile. Based on the comparison, one or more areas indicative of a height variation between the textile and the reference textile may be determined. An action may be performed based on the one or more areas indicative of the height variation.

Systems and methods for analyzing a fabric article to determine fiber damage are provided herein. In an embodiment of a method for analyzing a fabric article, the fabric article is provided and at least one image of the fabric article is captured using a three-dimensional imaging device. The at least one image is processed using a contrast function to produce a processed data set that is representative of fiber damage in a viewing area of the three-dimensional imaging device. Optionally, a processed image is produced based upon the processed data set.

A method of and device for testing and a scale for evaluating a textile for use in an article of clothing constructed of the textile. The method comprises positioning a textile sample within a path of a generated beam of light and measuring the amount of photons absorbed when the light is transmitted through the sample. The amount of light transmitted through the sample is compared to a standardized scale for evaluating the amount of coverage provided by the sample for the textiles fitness for use in the article of clothing. The articles of clothing may be articles of swimwear, athletic wear, or athleisure wear.

Computing devices and methods utilizing a joint speaker location/speaker identification neural network are provided. In one example a computing device receives an audio signal of utterances spoken by multiple persons. Magnitude and phase information features are extracted from the signal and inputted into a joint speaker location and speaker identification neural network. The neural network utilizes both the magnitude and phase information features to determine a change in the person speaking. Output comprising the determination of the change is received from the neural network. The output is then used to perform a speaker recognition function, speaker location function, or both.

Disclosed are a method and system for automatic detection of a fabric defect. The method comprises: preprocessing a region of a fabric to be detected during a fabric transmission process by a noise removing device to eliminate noise interference during a detection process; triggering a camera to perform image acquisition on the fabric to be detected according to rotation information of a motor of a roller of a cloth inspection machine; and automatically identifying a defect on the fabric to be detected based on an acquisition image.

An optical system includes an illumination module configured to illuminate a sample object with at least one angle-variable illumination geometry. The optical system includes an imaging optical unit configured to produce an imaged representation of the sample object that is illuminated with the at least one angle-variable illumination geometry on a detector. The optical system includes the detector, which is configured to capture at least one image of the sample object based on the imaged representation. The optical system includes a controller configured to determine a result image based on a transfer function and the at least one image. A method includes illuminating a sample object with at least one angle-variable illumination geometry, imaging the sample object on a detector, based on the imaged representation, capturing at least one image of the sample object, and, based on a transfer function and the at least one image, determining a result image.

A flaw detecting apparatus and a method for a plane mirror based on line scanning and ring band stitching are provided. The flaw detecting apparatus comprises: a line scanning detector, an annular illumination source, a rotary table rotatable about a Z axis, a translation table translatable along an X axis and a processor. By translating and rotating the plane mirror to be detected, an entire surface of the plane mirror to be detected can be detected by the line scanning detector, and the flaw of the entire plane mirror to be detected is obtained by a ring band stitching method. The method of line scanning and ring band stitching reduces the imaging distortion, the intermediate data amount, the difficulty in the distortion correction and difficulty in stitching, and improves the detection speed and the detection quality.

A real-time, full web image processing method for analyzing formation in a web is described, the web is transported in a moving direction during a web manufacturing process, the method including the steps of acquiring a two-dimensional original image P.sub.0 of the web, the image being representable as a digital image representable by a plurality of pixel values P.sub.0,i,j with i{1; . . . ; I}, j{1; . . . ; J}; and producing a plurality of P processed images P.sub.p with p{1; . . . ; P}, each of the processed images being representable by pixel values P.sub.p,m,n with m{1; . . . ; M}, n{1; . . . ; N}, the processed images being obtained by spatial bandpass filtering of the original image, wherein a spatial different bandpass filter is used for obtaining each of the processed.

Computing devices and methods utilizing a joint speaker location/speaker identification neural network are provided. In one example a computing device receives a multi-channel audio signal of an utterance spoken by a user. Magnitude and phase information features are extracted from the signal and inputted into a joint speaker location/speaker identification neural network that is trained via utterances from a plurality of persons. A user embedding comprising speaker identification characteristics and location characteristics is received from the neural network and compared to a plurality of enrollment embeddings extracted from the plurality of utterances that are each associated with an identity of a corresponding person. Based at least on the comparisons, the user is matched to an identity of one of the persons, and the identity of the person is outputted.

A surface defect inspection method includes: acquiring an original image by capturing an image of a subject of an inspection; generating texture feature images by applying a filtering process using spatial filters to the original image; generating a feature vector at each position of the original image, by extracting a value at a corresponding position from each of the texture feature images, for each of the positions of the original image; generating an abnormality level image representing an abnormality level for each position of the original image, by calculating, for each of the feature vectors, an abnormality level in a multi-dimensional distribution formed by the feature vectors; and detecting a part having the abnormality level that is higher than a predetermined level in the abnormality level image as a defect portion or a defect candidate portion.