Abstract: The described aspects and implementations enable efficient detection and classification of objects with machine learning models that deploy a bird's-eye view representation and are trained using depth ground truth data. In one implementation, disclosed are system and techniques that include obtaining images, generating, using a first neural network (NN), feature vectors (FVs) and depth distributions pixels of images, wherein the first NN is trained using training images and a depth ground truth data for the training images. The techniques further include obtaining a feature tensor (FT) in view of the FVs and the depth distributions, and processing the obtained FTs, using a second NN, to identify one or more objects depicted in the images.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for training a neural network to predict optical flow. One of the methods includes obtaining a batch of one or more training image pairs; for each of the pairs: processing the first training image and the second training image using the neural network to generate a final optical flow estimate; generating a cropped final optical flow estimate from the final optical flow estimate; and training the neural network using the cropped optical flow estimate.

Abstract: Disclosed herein are a surgical system for patella tracking and a method for selecting a properly-sized patellar implant utilizing the same. The surgical system may include first and second trackers and a patellar tracking system. The first tracker may be configured to contact an unresected or a resected patella, and the second tracker may be configured to contact a bone. The patellar tracking system may be configured to track the first and second trackers during patellar flexion and extension to generate patellar range of motion and patellar trial range of motion. A method for selecting a patellar implant may utilize the first and second trackers and the patellar tracking system.

Abstract: Disclosed herein are a surgical system for patella tracking and a method for selecting a properly-sized patellar implant utilizing the same. The surgical system may include first and second trackers and a patellar tracking system. The first tracker may be configured to contact an unresected or a resected patella, and the second tracker may be configured to contact a bone. The patellar tracking system may be configured to track the first and second trackers during patellar flexion and extension to generate patellar range of motion and patellar trial range of motion. A method for selecting a patellar implant may utilize the first and second trackers and the patellar tracking system.

Abstract: The described aspects and implementations enable efficient detection and classification of objects with machine learning models that deploy a bird's-eye view representation and are trained using depth ground truth data. In one implementation, disclosed are system and techniques that include obtaining images, generating, using a first neural network (NN), feature vectors (FVs) and depth distributions pixels of images, wherein the first NN is trained using training images and a depth ground truth data for the training images. The techniques further include obtaining a feature tensor (FT) in view of the FVs and the depth distributions, and processing the obtained FTs, using a second NN, to identify one or more objects depicted in the images.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for generating motion detection based on optical flow. One of the methods includes obtaining a first image of a scene in an environment taken by an agent at a first time point and a second image of the scene at a second later time point. A point cloud characterizing the scene in the environment is obtained. A predicted optical flow is determined between the first image and the second image. A respective initial flow prediction for the point that represents motion of the point between the two time points is determined. A respective ego motion flow estimate for the point that represents a motion of the point induced by ego motion of the agent is determined. A respective motion prediction that indicates whether the point was static or in motion between the two time points is determined.

Abstract: Disclosed herein are a surgical system for patella tracking and a method for selecting a properly-sized patellar implant utilizing the same. The surgical system may include first and second trackers and a patellar tracking system. The first tracker may be configured to contact an unresected or a resected patella, and the second tracker may be configured to contact a bone. The patellar tracking system may be configured to track the first and second trackers during patellar flexion and extension to generate patellar range of motion and patellar trial range of motion. A method for selecting a patellar implant may utilize the first and second trackers and the patellar tracking system.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for generating motion detection based on optical flow. One of the methods includes obtaining a first image of a scene in an environment taken by an agent at a first time point and a second image of the scene at a second later time point. A point cloud characterizing the scene in the environment is obtained. A predicted optical flow is determined between the first image and the second image. A respective initial flow prediction for the point that represents motion of the point between the two time points is determined. A respective ego motion flow estimate for the point that represents a motion of the point induced by ego motion of the agent is determined. A respective motion prediction that indicates whether the point was static or in motion between the two time points is determined.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for generating an optical flow label from a lidar point cloud. One of the methods includes obtaining data specifying a training example, including a first image of a scene in an environment captured at a first time point and a second image of the scene in the environment captured at a second time point. For each of a plurality of lidar points, a respective second corresponding pixel in the second image is obtained and a respective velocity estimate for the lidar point at the second time point is obtained. A respective first corresponding pixel in the first image is determined using the velocity estimate for the lidar point. A proxy optical flow ground truth for the training example is generated based on an estimate of optical flow of the pixel between the first and second images.

Abstract: Disclosed herein are a surgical system for patella tracking and a method for selecting a properly-sized patellar implant utilizing the same. The surgical system may include first and second trackers and a patellar tracking system. The first tracker may be configured to contact an unresected or a resected patella, and the second tracker may be configured to contact a bone. The patellar tracking system may be configured to track the first and second trackers during patellar flexion and extension to generate patellar range of motion and patellar trial range of motion. A method for selecting a patellar implant may utilize the first and second trackers and the patellar tracking system.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for training a neural network to predict optical flow. One of the methods includes obtaining a batch of one or more training image pairs; for each of the pairs: processing the first training image and the second training image using the neural network to generate a final optical flow estimate; generating a cropped final optical flow estimate from the final optical flow estimate; and training the neural network using the cropped optical flow estimate.

Abstract: A self-broaching, neck-preserving hip stem includes a set of rasp teeth on a distal-lateral portion of the stem to allow the stem to be impacted into a femur without the use of surgical broaches. The rasp teeth are disposed on a relief surface. The stem includes a length, radius of curvature, and a dual-taper in the anterior-posterior and distal-lateral directions which allow the neck to self-broach and seat itself proximal to the conventional resection level to preserve greater bone stock in the femur.

Abstract: Method and system of calibrating a sensor system which comprises sensor A and sensor B. Sensor A has a transmitter TXA for emitting a signal STXA and a receiver RXA for receiving a signal SRXA, wherein RXA and TXA operate independently in a radar mode of sensor A. Sensor B has a transmitter TXB, a receiver RXB, and a unit D, which is used to connect TXB to RXB in a transponder mode of sensor B, with the result that a signal SRXB received by RXB is emitted again by TXB as a signal STXB. A gain Gcon,B between signal SRXB and signal STXB is predefined. In a radar mode of sensor B, TXB is not connected to RXB, with the result that TXB and RXB operate independently. Emitted signals may be radar, light, or acoustic signals. The method and system can calibrate radar systems, lidar systems, or sonar systems.

Abstract: Disclosed herein are a surgical system for patella tracking and a method for selecting a properly-sized patellar implant utilizing the same. The surgical system may include first and second trackers and a patellar tracking system. The first tracker may be configured to contact an unresected or a resected patella, and the second tracker may be configured to contact a bone. The patellar tracking system may be configured to track the first and second trackers during patellar flexion and extension to generate patellar range of motion and patellar trial range of motion. A method for selecting a patellar implant may utilize the first and second trackers and the patellar tracking system.

Abstract: The invention relates to a method for calibrating an active sensor system which comprises at least one active sensor A and one active sensor B. The term “active” indicates that the sensors A and B actively emit signals. The sensor A has a transmitter TXA for emitting a signal STXA and a receiver RXA for receiving a signal S RXA, wherein the receiver RX A and the transmitter TXA operate independently of one another to the greatest possible extent in a radar mode of the sensor A. The sensor B has a transmitter TXB, a receiver RXB, and a unit D, which is used to connect the transmitter TXB to the receiver RXB in a transponder mode of the sensor B, with the result that a signal SRXB received by the receiver RXB is emitted again by the transmitter TXB as a signal STXB. A gain Gcon,B between the received signal SRXB and the signal STXB which is emitted again is predefined in this case.

Abstract: The invention provides a heavy duty rim assembly (10) for use with a moulded or cast non-inflatable tyre, a wheel (100) including such a rim assembly and tyre (102), a mould (200) for such a tyre, and a method of manufacturing such a wheel. The rim assembly includes a nave (12) for mounting on a vehicle axle, a hollow barrel (14) around the nave, and a plurality of key members (16) around the barrel. The key members, in use, provide for effective bonding and keying of the rim assembly with the non-inflatable tyre. The key members define a belt (46) around the barrel, the belt being at least substantially concentric with and spaced apart from the barrel. The key members define holes (36) therethrough for tyre material to extend through in a completed wheel.

Abstract: Traps particularly intended for carrying condensate from air conditioner drip trays towards a drainage system are constructed from a single turn of a flexible thermoplastic pipe; each unit being preformed when in a heat-softened state into a shape of a one-turn coil resembling the twist in a pig's tail, or the Greek “alpha” (&agr;) symbol. Traps are made of a transparent flexible durable thermoplastics material, such as polyvinyl chloride, and internal contents may be inspected through the trap wall. Being flexible, the traps can reliably connect mis-aligned conduits. Being flexible, the traps can serve as reliable connectors between conduits where relative movement is a problem such as where vibration or earthquakes occur.

Abstract: Traps particularly intended for carrying condensate from air conditioner drip trays towards a drainage system are constructed from a single turn of a flexible thermoplastic pipe; each unit being preformed when in a heat-softened state into a shape of a one-turn coil resembling the twist in a pig's tail, or the Greek “alpha” (&agr;) symbol. Traps are made of a transparent flexible durable thermoplastics material, such as polyvinyl chloride, and internal contents may be inspected through the trap wall. Being flexible, the traps can reliably connect mis-aligned conduits. Being flexible, the traps can serve as reliable connectors between conduits where relative movement is a problem such as where vibration or earthquakes occur.