Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media for identification and re-identification of fish. In some implementations, first media representative of aquatic cargo is received. Second media based on the first media is generated, wherein a resolution of the second media is higher than a resolution of the first media. A cropped representation of the second media is generated. The cropped representation is provided to the machine learning model. In response to providing the cropped representation to the machine learning model, an embedding representing the cropped representation is generated using the machine learning model. The embedding is mapped to a high dimensional space. Data identifying the aquatic cargo is provided to a database, wherein the data identifying the aquatic cargo comprises an identifier of the aquatic cargo, the embedding, and a mapped region of the high dimensional space.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer-storage media, for controlling an aquaculture feeding system are described. In some implementations, a method includes determining, using a feeding system controller, that a feeding system has transitioned from a first state to a second state, and based on the transition from the first to the second state providing, using the feeding system controller, instructions to a dosing system to set a feed rate of the dosing system, and providing, using the feeding system controller, instructions to a blower operatively coupled to the dosing system to set a flow rate of the blower.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer-storage media, for fish weight estimation based on fish tracks identified in images. In some implementations, a method includes obtaining images of fish enclosed in a fish enclosure, identifying fish tracks shown in the images of the fish, determining a quality score for each of the fish tracks, selecting a subset of the fish tracks based on the quality scores, determining a representative weight of the fish in the fish enclosure based on weights of the fish shown in the subset of the fish tracks, and outputting the representative weight for display or storage at a device connected to the one or more processors.

Abstract: A method for controlling a sensor subsystem, the method including receiving one or more metrics representing one or more characteristics of livestock, including one or more livestock objects, contained in an enclosure and monitored by one or more sensors coupled to a winch subsystem. The method further includes determining a position to move the one or more sensors based on the metrics and determining an instruction that includes information related to a movement of the one or more sensors. The method further includes sending the instruction to the winch subsystem to change the position of the one or more sensors.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer-storage media, for controlling a camera to observe aquaculture feeding behavior. In some implementations, a method includes moving a camera to a first position, obtaining an image captured by the camera at the first position, determining a feeding observation mode, and based on the feeding observation mode and analysis of the image, determining a second position to move the camera.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, that control devices in an aquaculture environment. One of the methods includes determining a particular objective for a robot that is operating in an aquaculture environment and determining one or more sensed conditions that are associated with the aquaculture environment. The particular objective is provided to an anti-fish-startling model evaluation engine that is configured to output actions, for a given objective, that accomplish the given objective while reducing a startling effect on nearby fish. Based on providing the particular objective to the anti-fish-startling model evaluation engine, one or more particular actions for accomplishing the particular objective are determined. The one or more particular actions are transmitted to another device.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium that automatically performs actions in an aquaculture environment based on light sensed by underwater cameras. One of the methods includes obtaining images of a surface of water captured by a camera that faces upwards from a depth towards the surface of the water within an enclosure that encloses aquatic livestock. An ambient light metric is determined at the depth from the images of the surface of the water. A determination is made as to whether the camera satisfies one or more depth criteria. Based on determining that the depth of the camera satisfies the one or more depth criteria, it is determined that, based on the ambient light metric at the depth, one or more action criteria are satisfied, then initiating performance of an action to be performed for the aquatic livestock.

Abstract: Methods, systems, and apparatuses, including computer programs encoded on a computer-readable storage medium for estimating the shape, size, and mass of fish are described. A pair of stereo cameras may be utilized to obtain right and left images of fish in a defined area. The right and left images may be processed, enhanced, and combined. Object detection may be used to detect and track a fish in images. A pose estimator may be used to determine key points and features of the detected fish. Based on the key points, a three-dimensional (3-D) model of the fish is generated that provides an estimate of the size and shape of the fish. A regression model or neural network model can be applied to the 3-D model to determine a likely weight of the fish.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer-storage media, for sorting fish in aquaculture. In some implementations, one or more images are obtained of a particular fish within a population of fish. Based on the one or more images of the fish, a data element is determined. The data element can include a first value that reflects a physical characteristic of the particular fish, and a second value that reflects a condition factor of the particular fish. Based on the data element, the fish is classified as a member of a particular subpopulation of the population of fish. An actuator of an automated fish sorter is controlled based on classifying the particular fish as a member of the particular subpopulation of the population of fish.

Abstract: In one aspect, there is provided a method that includes receiving, by a control system having (i) a first camera configured to obtain an image of a scene, (ii) a winch controller, and (iii) a feeding system configured to deliver a feed to aquaculture, instructions to initiate a calibration of the first camera, determining a calibration state of the first camera, determining a sequence of calibration steps based on the calibration state of the first camera, and executing the sequence of calibration steps to calibrate the first camera.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media for identification and re-identification of fish. In some implementations, first media representative of aquatic cargo is received. Second media based on the first media is generated, wherein a resolution of the second media is higher than a resolution of the first media. A cropped representation of the second media is generated. The cropped representation is provided to the machine learning model. In response to providing the cropped representation to the machine learning model, an embedding representing the cropped representation is generated using the machine learning model. The embedding is mapped to a high dimensional space. Data identifying the aquatic cargo is provided to a database, wherein the data identifying the aquatic cargo comprises an identifier of the aquatic cargo, the embedding, and a mapped region of the high dimensional space.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, that generate from a first pair and a second pair of images of livestock that are within an enclosure and that are taken at different times using a stereoscopic camera, at least two distance distributions of the aquatic livestock within the enclosure. The distance distributions can be used to determine a measure associated with an optical property of the water within the enclosure. A signal associated with the measure can be provided.

Abstract: Methods, systems, and apparatus for receiving, by a voice action system, data specifying trigger terms that trigger an application to perform a voice action and a context that specifies a status of the application when the voice action can be triggered. The voice action system receives data defining a discoverability example for the voice action that comprises one or more of the trigger terms that trigger the application to perform the voice action when a status of the application satisfies the specified context. The voice action system receives a request for discoverability examples for the application from a user device having the application installed, and provides the data defining the discoverability examples to the user device in response to the request. The user device is configured to provide a notification of the one or more of the trigger terms when a status of the application satisfies the specified context.

Abstract: A method for controlling a sensor subsystem, the method including receiving one or more metrics representing one or more characteristics of livestock, including one or more livestock objects, contained in an enclosure and monitored by one or more sensors coupled to a winch subsystem. The method further includes determining a position to move the one or more sensors based on the metrics and determining an instruction that includes information related to a movement of the one or more sensors. The method further includes sending the instruction to the winch subsystem to change the position of the one or more sensors.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer-storage media, for controlling an aquaculture feeding system are described. In some implementations, a method includes determining, using a feeding system controller, that a feeding system has transitioned from a first state to a second state, and based on the transition from the first to the second state providing, using the feeding system controller, instructions to a dosing system to set a feed rate of the dosing system, and providing, using the feeding system controller, instructions to a blower operatively coupled to the dosing system to set a flow rate of the blower.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer-readable storage media, for automated camera positioning for feeding behavior monitoring. In some implementations, a system obtains an image of a scene, a spatial model that corresponds to a subfeeder, and calibration parameters of a camera, the system determines a size of the subfeeder in the image of the scene, the system selects an updated position of the camera relative to the subfeeder, the system provides the updated position of the camera relative to the subfeeder to a winch controller, and the system moves the camera to the updated position.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer-storage media, for controlling a camera to observe aquaculture feeding behavior. In some implementations, a method includes moving a camera to a first position, obtaining an image captured by the camera at the first position, determining a feeding observation mode, and based on the feeding observation mode and analysis of the image, determining a second position to move the camera.

Abstract: A fish monitoring system deployed in a particular area to obtain fish images is described. Neural networks and machine-learning techniques may be implemented to periodically train fish monitoring systems and generate monitoring modes to capture high quality images of fish based on the conditions in the determined area. The camera systems may be configured according to the settings, e.g., positions, viewing angles, specified by the monitoring modes when conditions matching the monitoring modes are detected. Each monitoring mode may be associated with one or more fish activities, such as sleeping, eating, swimming alone, and one or more parameters, such as time, location, and fish type.

Abstract: Methods, systems, and apparatuses, including computer programs encoded on a computer-readable storage medium for estimating the shape, size, mass, and health of fish are described. A pair of stereo cameras may be utilized to obtain off-axis images of fish in a defined area. The images may be processed, enhanced, and combined. Object detection may be used to detect and track a fish in images. A pose estimator may be used to determine key points and features of the detected fish. Based on the key points, a model of the fish is generated that provides an estimate of the size and shape of the fish. A regression model or neural network model can be applied to the fish model to determine characteristics of the fish.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer-storage media, for escape detection and mitigation for aquaculture. In some implementations, a method includes obtaining one or more images that depict one or more fish within a population of fish that are located within an enclosure; providing, to one or more detection models configured to classify fish that are depicted within the images as likely being member or as likely not being member of a type of fish, the one or images; generating, as a result of providing the one or more images to the one or more detection models, a value that reflects a quantity of fish that are depicted in the images that are likely a member of the type of fish; and detecting a condition based at least on the value.