Abstract: Generation of high-performance machine learning models often requires significant computational resources and access to extensive training datasets. This makes development of such models for rare or novel diseases, where diagnostic imagery or other training data is limited, difficult. Methods are provided to apply extensive generic medical imagery training datasets to train machine learning models to embed input medical imaging data into generically informative embedding spaces. Relatively smaller training datasets specific to a novel or rare disease can then be used to develop high-performance models by updating the parameters of the pre-trained generic model and/or by training a smaller, task-specific model to predict one or more variables of interest based on embedding vectors output from the pre-trained generic model. The functionality of such a generic model can be made available via an online service to facilitate development of such task-specific models by smaller research groups.

Abstract: The present disclosure provides systems and methods that perform face reconstruction based on an image of a face. In particular, one example system of the present disclosure combines a machine-learned image recognition model with a face modeler that uses a morphable model of a human's facial appearance. The image recognition model can be a deep learning model that generates an embedding in response to receipt of an image (e.g., an uncontrolled image of a face). The example system can further include a small, lightweight, translation model structurally positioned between the image recognition model and the face modeler. The translation model can be a machine-learned model that is trained to receive the embedding generated by the image recognition model and, in response, output a plurality of facial modeling parameter values usable by the face modeler to generate a model of the face.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for generating realistic extensions of images. In one aspect, a method comprises providing an input that comprises a provided image to a generative neural network having a plurality of generative neural network parameters. The generative neural network processes the input in accordance with trained values of the plurality of generative neural network parameters to generate an extended image. The extended image has (i) more rows, more columns, or both than the provided image, and (ii) is predicted to be a realistic extension of the provided image. The generative neural network is trained using an adversarial loss objective function.

Abstract: Techniques are provided for input/output operations per second (IOPS) and throughput monitoring for dynamic and/or optimal resource allocation. These techniques provide automated monitoring of resources, such as memory and processor utilization by a container accessing a volume. The automated monitoring is performed in order to generate and execute intelligent recommendations for improved resource utilization. Resource allocations can be scaled up to meet I/O load demand and satisfy service level agreements (SLAs). Resource allocations can be scaled down or adjusted to conserve resources, such as by consolidating containers or pods hosted in multiple virtual machines into a single virtual machine and decommissioning virtual machines no longer hosting containers or pods.

Abstract: Systems and methods for generating insect classifications using predictive models based on sequences of images are disclosed. An example system includes an imaging device configured to capture images of insects and a computing device in communication with the imaging device. The computing device is configured to instruct the imaging device to capture a sequence of images depicting at least a portion of an insect. The computing device is further configured to use a first predictive model to determine a first output corresponding to a first classification of a first image of the sequence of images, the first output including a confidence measure of the first classification. The computing device is further configured to generate classification information based at least in part on the first output.

Abstract: To adjust an aspect ratio of an image to match the aspect ratio of a display area for presenting the image, a computing device receives an image having a first aspect ratio, and obtains a second aspect ratio for a display area of a display in which to present the image, where the second aspect ratio is different from the first aspect ratio. The computing device extends the image to include one or more additional features which were not included in the image. Additionally, the computing device automatically crops the extended image around an identified region of interest by selecting a portion of the extended image that has an aspect ratio which matches the second aspect ratio of the display area, and provides the cropped image for presentation within the display area of the display.

Abstract: Techniques are provided for input/output operations per second (IOPS) and throughput monitoring for dynamic and/or optimal resource allocation. These techniques provide automated monitoring of resources, such as memory and processor utilization by a container accessing a volume. The automated monitoring is performed in order to generate and execute intelligent recommendations for improved resource utilization. Resource allocations can be scaled up to meet I/O load demand and satisfy service level agreements (SLAs). Resource allocations can be scaled down or adjusted to conserve resources, such as by consolidating containers or pods hosted in multiple virtual machines into a single virtual machine and decommissioning virtual machines no longer hosting containers or pods.

Abstract: An insect sortation system is disclosed. The insect sortation system includes a pathway having a first end and a second end; a first outlet extending from the pathway adjacent to the second end; a second outlet extending from the pathway adjacent to the second end; a detector disposed adjacent to the pathway and configured to detect one or more characteristics of mosquitos passing by the detector; a shutter disposed within the pathway between the detector and the second end of the pathway; and one or more sortation devices configured to direct each mosquito of the mosquitos to at least one of the first outlet or the second outlet after passing by the detector.

Abstract: Insects can be classified into a category (e.g., sex category, species category, size category, etc.) using a variety of different classification approaches including, for example, an industrial vision classifier and/or a machine learning classifier. At least some classification approaches may be used in real-time to make real-time decisions and others can be used to validate earlier-made real-time decisions.

Abstract: In some implementations, a device may generate, based at least in part on a first set of inputs, a serverless software development environment associated with a set of cloud resources. The device may generate, based at least in part on a first machine learning model, a technology stack recommendation having a set of associated tools for performing a software development task. The device may instantiate the selected technology stack in the serverless software development environment and generate a set of applications based at least in part on executing the set of tools. The device may deploy the set of applications in one or more serverless application environments. The device may use machine learning to observe deployed applications, detect hidden anomalies, and perform root-cause analysis, thereby providing a lean and sustainable serverless environment.

Abstract: In some implementations, a device may generate, based at least in part on a first set of inputs, a serverless software development environment associated with a set of cloud resources. The device may generate, based at least in part on a first machine learning model, a technology stack recommendation having a set of associated tools for performing a software development task. The device may instantiate the selected technology stack in the serverless software development environment and generate a set of applications based at least in part on executing the set of tools. The device may deploy the set of applications in one or more serverless application environments. The device may use machine learning to observe deployed applications, detect hidden anomalies, and perform root-cause analysis, thereby providing a lean and sustainable serverless environment.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for training a neural network. One of the methods includes processing each training input using the neural network and in accordance with the current values of the network parameters to generate a network output for the training input; computing a respective loss for each of the training inputs by evaluating a loss function; identifying, from a plurality of possible perturbations, a maximally non-linear perturbation; and determining an update to the current values of the parameters of the neural network by performing an iteration of a neural network training procedure to decrease the respective losses for the training inputs and to decrease the non-linearity of the loss function for the identified maximally non-linear perturbation.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for training a neural network. One of the methods includes processing each training input using the neural network and in accordance with the current values of the network parameters to generate a network output for the training input; computing a respective loss for each of the training inputs by evaluating a loss function; identifying, from a plurality of possible perturbations, a maximally non-linear perturbation; and determining an update to the current values of the parameters of the neural network by performing an iteration of a neural network training procedure to decrease the respective losses for the training inputs and to decrease the non-linearity of the loss function for the identified maximally non-linear perturbation.

Abstract: Techniques are provided for input/output operations per second (lOPS) and throughput monitoring for dynamic and/or optimal resource allocation. These techniques provide automated monitoring of resources, such as memory and processor utilization by a container accessing a volume. The automated monitoring is performed in order to generate and execute intelligent recommendations for improved resource utilization. Resource allocations can be scaled up to meet l/O load demand and satisfy service level agreements (SLAs). Resource allocations can be scaled down or adjusted to conserve resources, such as by consolidating containers or pods hosted in multiple virtual machines into a single virtual machine and decommissioning virtual machines no longer hosting containers or pods.

Abstract: The present disclosure provides an improved training methodology that enables supervised contrastive learning to be simultaneously performed across multiple positive and negative training examples. In particular, example aspects of the present disclosure are directed to an improved, supervised version of the batch contrastive loss, which has been shown to be very effective at learning powerful representations in the self-supervised setting Thus, the proposed techniques adapt contrastive learning to the fully supervised setting and also enable learning to occur simultaneously across multiple positive examples.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for training a neural network. One of the methods includes processing each training input using the neural network and in accordance with the current values of the network parameters to generate a network output for the training input; computing a respective loss for each of the training inputs by evaluating a loss function; identifying, from a plurality of possible perturbations, a maximally non-linear perturbation; and determining an update to the current values of the parameters of the neural network by performing an iteration of a neural network training procedure to decrease the respective losses for the training inputs and to decrease the non-linearity of the loss function for the identified maximally non-linear perturbation.

Abstract: The present disclosure provides systems and methods that perform face reconstruction based on an image of a face. In particular, one example system of the present disclosure combines a machine-learned image recognition model with a face modeler that uses a morphable model of a human's facial appearance. The image recognition model can be a deep learning model that generates an embedding in response to receipt of an image (e.g., an uncontrolled image of a face). The example system can further include a small, lightweight, translation model structurally positioned between the image recognition model and the face modeler. The translation model can be a machine-learned model that is trained to receive the embedding generated by the image recognition model and, in response, output a plurality of facial modeling parameter values usable by the face modeler to generate a model of the face.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for processing images using an image processing neural network system. One of the system includes a shared encoder neural network implemented by one or more computers, wherein the shared encoder neural network is configured to: receive an input image from a target domain; and process the input image to generate a shared feature representation of features of the input image that are shared between images from the target domain and images from a source domain different from the target domain; and a classifier neural network implemented by the one or more computers, wherein the classifier neural network is configured to: receive the shared feature representation; and process the shared feature representation to generate a network output for the input image that characterizes the input image.

Abstract: The present disclosure provides an improved training methodology that enables supervised contrastive learning to be simultaneously performed across multiple positive and negative training examples. In particular, example aspects of the present disclosure are directed to an improved, supervised version of the batch contrastive loss, which has been shown to be very effective at learning powerful representations in the self-supervised setting. Thus, the proposed techniques adapt contrastive learning to the fully supervised setting and also enable learning to occur simultaneously across multiple positive examples.

Abstract: The present disclosure provides systems and methods that perform face reconstruction based on an image of a face. In particular, one example system of the present disclosure combines a machine-learned image recognition model with a face modeler that uses a morphable model of a human's facial appearance. The image recognition model can be a deep learning model that generates an embedding in response to receipt of an image (e.g., an uncontrolled image of a face). The example system can further include a small, lightweight, translation model structurally positioned between the image recognition model and the face modeler. The translation model can be a machine-learned model that is trained to receive the embedding generated by the image recognition model and, in response, output a plurality of facial modeling parameter values usable by the face modeler to generate a model of the face.