Abstract: The present disclosure provides a wireless receiver having a Delay Doppler channel detector. The Delay Doppler channel detector includes a Delay Dopper domain channel estimator and a detector and demodulator. The Delay Doppler domain channel estimator is configured for estimating a channel gain, a delay, and a Doppler frequency shift of a path of a channel of data received via the channel. The detector and demodulator is configured for extracting a symbol sequence of the received data based on the estimated channel gain, the estimated delay, and the estimated Doppler frequency shift of the path of the channel.

Abstract: A method for performing a crop yield estimation using a semi-supervised deep convolution neural network is provided. The method includes receiving monitoring data from a drone, wherein the monitoring data comprises a video of the crops captured by the drone; sampling the video by a predefined frame rate to obtain one or more images; inputting the images to a crop yield estimation model to obtain one or more result data, wherein the crop yield estimation model comprises a generator and a discriminator each comprising one or more DCNNs, and wherein the crop yield estimation model is trained by a semi-supervised learning method; and performing a quantity estimation and a quality estimation corresponding to the crops as shown in the images according to the one or more result data, so as to determine a total number and maturities of the crops respectively.

Abstract: An integrated MWP processing engine based on a thin-film lithium niobate (LN) platform capable of performing computation tasks of analog signals up to 92G samples per second is provided. By integrating a high-speed electro-optic modulation block and a multi-purpose low-loss signal processing section on the same chip fabricated from a 4-inch wafer-scale process, we demonstrate high-speed analog computation, i.e. first- and second-order temporal integration and differentiation with computing accuracies up to 98.1%, and deploy these functions to showcase three proof-of-concept applications, namely, ordinary differential equation solving, ultra-wideband signal generation and high-speed edge detection of images. We further leverage the image edge detector to enable a photonic-assisted image segmentation model that effectively outlines boundaries of melanoma lesion in medical images, achieving orders of magnitude faster processing speed and lower power consumption than conventional electronic processors.

Abstract: The present invention provides an unsupervised domain adaptive segmentation network comprises a feature extractor configured for extracting features from a 3D MRI scan image; a decorrelation and whitening module configured for preforming decorrelation and whitening transformation on the extracted features to obtain whitened features; a domain-specific feature translation module configured for translating domain-specific features from a source domain into a target domain for adapting the unsupervised domain adaptive network to the target domain; and a classifier configured for projecting the whitened features into a zonal segmentation prediction. By implementing the domain-specific feature translation module for transferring the knowledge learned from the labeled source domain data to unlabeled target domain data, domain gap between the source and target data can be narrowed.

Abstract: A system and method for generating a stained image including the steps of obtaining a first image of a key sample section; and processing the first image with a multi-modal stain learning engine arranged to generate at least one stained image, wherein the at least one stained image represents the key sample section stained with at least one stain.

Abstract: The present invention provides a graph-based prostate diagnosis network (GPD-Net) and a method for using the same to predict a prostate health status of a patient from a 3D magnetic resonance imaging (MRI) scan containing a plurality of 2D MRI slices. The GPD-Net only demands patient-level annotations of MRI scan for training by formulating the diagnosis task of 3D prostate MRI scan in a multi-instance learning (MIL) strategy, and regarding each 2D MRI slice in the 3D prostate MRI scan as an instance. The GPD-Net includes a plurality of importance-guided graph convolutional layers to explore the diagnostic information with the importance-based topology. The present invention provides accurate prediction of prostate diseases and achieve more reliable diagnosis from MRI scans, therefore can effectively alleviate the workload of clinician in viewing the slices of MRI scan.

Abstract: The present invention provides a task interaction network which can jointly perform, based on multi parametric-magnetic resonance imaging scan images, a segmentation task to locate prostate cancer areas and a classification task to access aggressiveness of lesions. The task interaction network comprises a backbone network, an auxiliary segmentation branch, a classification branch having a lesion awareness module, and a main segmentation branch having a category allocation module. The auxiliary segmentation branch is utilized to predict an initial lesion mask as location guidance information for the classification branch to perform the classification task. The lesion awareness module is configured to refine the initial lesion mask to make it more accurate. Moreover, weights used in classification branch can serve as the category prototypes for generating category guidance features via the category allocation module to assist the main segmentation branch to perform the segmentation task.

Abstract: The present invention provides a graph-based prostate diagnosis network (GPD-Net) and a method for using the same to predict a prostate health status of a patient from a 3D magnetic resonance imaging (MRI) scan containing a plurality of 2D MRI slices. The GPD-Net only demands patient-level annotations of MRI scan for training by formulating the diagnosis task of 3D prostate MRI scan in a multi-instance learning (MIL) strategy, and regarding each 2D MRI slice in the 3D prostate MRI scan as an instance. The GPD-Net includes a plurality of importance-guided graph convolutional layers to explore the diagnostic information with the importance-based topology. The present invention provides accurate prediction of prostate diseases and achieve more reliable diagnosis from MRI scans, therefore can effectively alleviate the workload of clinician in viewing the slices of MRI scan.

Abstract: A method for performing a crop yield estimation using a semi-supervised deep convolution neural network is provided. The method includes receiving monitoring data from a drone, wherein the monitoring data comprises a video of the crops captured by the drone; sampling the video by a predefined frame rate to obtain one or more images; inputting the images to a crop yield estimation model to obtain one or more result data, wherein the crop yield estimation model comprises a generator and a discriminator each comprising one or more DCNNs, and wherein the crop yield estimation model is trained by a semi-supervised learning method; and performing a quantity estimation and a quality estimation corresponding to the crops as shown in the images according to the one or more result data, so as to determine a total number and maturities of the crops respectively.

Abstract: The present invention provides an unsupervised domain adaptive segmentation network comprises a feature extractor configured for extracting features from a 3D MRI scan image; a decorrelation and whitening module configured for preforming decorrelation and whitening transformation on the extracted features to obtain whitened features; a domain-specific feature translation module configured for translating domain-specific features from a source domain into a target domain for adapting the unsupervised domain adaptive network to the target domain; and a classifier configured for projecting the whitened features into a zonal segmentation prediction. By implementing the domain-specific feature translation module for transferring the knowledge learned from the labeled source domain data to unlabeled target domain data, domain gap between the source and target data can be narrowed.

Abstract: The present invention provides a task interaction network which can jointly perform, based on multi parametric-magnetic resonance imaging scan images, a segmentation task to locate prostate cancer areas and a classification task to access aggressiveness of lesions. The task interaction network comprises a backbone network, an auxiliary segmentation branch, a classification branch having a lesion awareness module, and a main segmentation branch having a category allocation module. The auxiliary segmentation branch is utilized to predict an initial lesion mask as location guidance information for the classification branch to perform the classification task. The lesion awareness module is configured to refine the initial lesion mask to make it more accurate. Moreover, weights used in classification branch can serve as the category prototypes for generating category guidance features via the category allocation module to assist the main segmentation branch to perform the segmentation task.

Abstract: A system and method for generating a stained image including the steps of obtaining a first image of a key sample section; and processing the first image with a multi-modal stain learning engine arranged to generate at least one stained image, wherein the at least one stained image represents the key sample section stained with at least one stain.

Abstract: A system and method for generating a stained image including the steps of obtaining a first image of a key sample section; and processing the first image with a stain learning engine arranged to generate at least one stained image, wherein the at least one stained image represents the key sample section stained with at least one stain.

Abstract: A system and method for generating a stained image including the steps of obtaining a first image of a key sample section; and processing the first image with a stain learning engine arranged to generate at least one stained image, wherein the at least one stained image represents the key sample section stained with at least one stain.