Abstract: An edge device includes a first antenna array, control circuitry, and a sensing radar communicatively coupled to the control circuitry, where the sensing radar senses a surrounding area of the edge device. The control circuitry concurrently communicates two different beams of radio frequency (RF) signals to two user equipment (UEs) from two portions of the first antenna array based on the sensed surrounding area of the edge device by the sensing radar. The sensing radar and the first antenna array employ mutually isolated frequencies for the sensing and the communication, respectively.

Abstract: Provided is an altitude calculation method for a flying body including calculating an altitude of the flying body by using information on horizontal coordinates of the flying body, a reception signal strength of a radio signal, and altitude calculation auxiliary information obtained in advance.

Abstract: Disclosed are techniques for identifying virtual anchors. In an aspect, a network entity obtains a set of data samples associated with a user equipment (UE), each data sample comprising a location of the UE and a set of multipath components (MPCs) obtained by the UE, and determines whether any MPCs in a subset of non-line-of-sight (NLOS) MPCs selected from at least one set of MPCs are associated with any previously identified virtual anchors in a database of found virtual anchors, wherein one or more first NLOS MPCs are removed from the set of data samples based on the one or more first NLOS MPCs being associated with a previously identified virtual anchor, and wherein one or more second NLOS MPCs are added to the database of found virtual anchors based on the one or more second NLOS MPCs not being associated with any previously identified virtual anchor.

Abstract: Methods, apparatus and systems for wireless human-nonhuman motion detection are described.

Abstract: In one aspect, a method of determining a location of a user within an indoor space, includes emitting a radiofrequency signal into the indoor space, receiving backscattered training radiofrequency signals, including multipath, for at least one location within the indoor space, converting the received training signals into a point cloud for each location of the at least one location, assigning a signature for each location based on the point cloud for each location, receiving additional radiofrequency signals, including multipath, converting the additional radiofrequency signals into an additional point cloud, and determining a location of the user by comparing the additional point cloud to the assigned signatures.

Abstract: The present disclosure relates to a method and device for communication by a base station in a wireless communication system supporting multiple reconfigurable intelligent surface (RIS) devices. According to an embodiment A method for operating a base station, the method comprises selecting at least two reference reconfigurable intelligent surface (RIS) devices for estimating a position of a user equipment (UE) from among a plurality of RIS devices, estimating position information about the UE using the selected at least two reference RIS devices, receiving an uplink pilot signal of the UE through the selected at least two reference RIS devices; and estimating a multi-RIS channel for the plurality of RIS devices based on the uplink pilot signal of the UE and the estimated position information about the UE.

Abstract: An interference signal detection method, apparatus, integrated circuit, radio device and terminal are disclosed, wherein, the method includes: transmitting a target signal based on a radar, receiving N frames of echo signals, obtaining N pieces of 2D FFT plane information according to the N frames of echo signals, wherein any piece of 2D FFT plane information includes power information corresponding to a range and a velocity, and detecting whether an echo signal is interfered according to accumulated power information corresponding to the range gate in the piece of 2D FFT plane information. In this way, the interference is characterized by the accumulated power information, which has good robustness and simple operations, and can achieve low power consumption of the radar system.

Abstract: Method for providing reduced interference for at least two co-located FMCW (Frequency Modulated Continuous Wave) Doppler radars, each of said radars being used in a system to detect respective distances to and velocities of objects moving through space, can include a propagation determination step, in which expected electromagnetic wave propagation times are determined between pairs of radars; a sweep time offset synchronizing step, in which different respective sweep time offsets are selected, with respect to each radar in a first group of radars; and a sweep frequency offset synchronizing step, in which a second sweep frequency offset is selected with respect to a second group of radars, the second sweep frequency offset being relative to a sweep frequency pattern used for radars belonging to said first group. The invention also relates to a system and to a computer software product.

Abstract: The present disclosure provides a radar apparatus capable of suppressing a radar-to-radar in-band interference caused by a signal derived from another radar apparatus in a radar system in which a plurality of radar apparatuses share frequency resource to increase a frequency utilization efficiency. The radar apparatus includes: a receiver configured to acquire a received signal; a plurality of matched filters arranged in parallel to filter the receive signal, each corresponding to respective one of the plurality of radar apparatus to use a transmit signal of a corresponding radar apparatus as a reference signal; and a signal reconstructor configured to select a matched filter most adequate for an interference suppression based on outputs of the plurality of matched filters, subtract the reference signal of a selected matched filter at least partially from the receive signal, and reconstruct a desired echo signal to be used to determine a target.

Abstract: The present subject matter relates to a radar system for detecting the surroundings of a moving object, in particular a vehicle and/or a transport device, such as in particular a crane, wherein the system is mounted or mountable on the moving object, wherein the radar system comprises at least one first, non-coherent, and at least one second, non-coherent, radar module with at least one antenna, wherein the radar modules are arranged or can be arranged distributed on the moving object, wherein at least one first radar module is configured differently from at least one second radar module.

Abstract: A pulse generator comprising: a first signal generating arm comprising a first inductor and a plurality of switching elements, each arranged to draw current through the first inductor; and a controller arranged to activate the plurality of switching elements in a predetermined sequence so as to generate a predetermined pulse waveform at a pulse generator output. The switching elements of the signal generating arm and the inductor together form a pulse synthesizer that takes the signal from the controller and uses it to synthesize an output pulse. Compared with conventional transmitter architectures, the functions of the upconversion mixer, the DAC, and the power amplifier are all performed by a single simplified circuit. This is both area efficient and power efficient.

Abstract: An apparatus for driver assistance includes a radar installed to a vehicle, having a sensing area outside the vehicle, and configured to provide object data, and a controller configured to identify a distance to an object from the vehicle and a moving speed of the object based on the object data. The radar includes an antenna array, a signal processing circuit configured to provide a transmission signal to the antenna array to transmit radio waves and acquire a reception signal corresponding to radio waves received by the antenna array, and a signal processor configured to control the signal processing circuit to provide a pre-chirp signal including a plurality of pre-chirps and a main chirp signal including a plurality of main chirps to the antenna array. The number of pre-chirps is less than the number of main chirps.

Abstract: An integrated circuit (IC) includes circuitry in a plurality of power domains for transmitting and/or receiving radar chirp frames and first and second monitoring systems for monitoring supply voltages of a first and a second subset of the plurality of power domains, respectively. The first subset is monitored outside of a time window during which a chirp frame is transmitted and/or received utilizing circuitry of the IC, and the second subset is monitored during the time window. The first monitoring system includes an output for an error signal indicating a supply voltage in the first subset does not comply with a first voltage parameter. The second monitoring system includes a unique monitoring circuit for each power domain in the second subset, and each unique monitoring circuit includes an output for an error signal indicating a supply voltage in the second subset does not comply with a second voltage parameter.

Abstract: A method of designing a radar, comprises receiving data pertaining to a set of reflected signals received from a distribution of objects by a respective set of receiving antennas at a respective set of locations, and feeding the data and the locations as training data to a machine learning procedure. The machine learning procedure calculates, simultaneously, a set of learned antenna locations and a set of learned parameters associating the signals with the objects, thereby providing a trained machine learning procedure parametrized by the set of learned parameters.

Abstract: Techniques for segmenting and classifying a representation of aggregated sensor data from a scene are discussed herein. Sensor data may be collected during multiple traversals of a same scene, and the sensor data may be filtered to remove portions of the sensor data not relevant to road network maps. In some examples, the filtered data may be aggregated and represented in voxels of a three-dimensional voxel space, from which an image representing a top-down view of the scene may be generated, though other views are also contemplated. Operations may include segmenting and/or classifying the image e.g., by a trained machine-learned model, to associate class labels indicative of map elements (e.g., driving lane, stop line, turn lane, and the like) with segments identified in the image. Additionally, techniques may create or update road network maps based on segmented and semantically labeled image(s) of various portions of an environment.

Abstract: A road surface condition determination system includes: a ranging sensor that is installed at a predetermined position on a vehicle, emits light toward a road surface in a measurement area around the vehicle and receives reflected light of the emitted light, and measures a distance to the road surface based on a light reception state of the reflected light; and a determination unit that determines that the road surface is wet when the reflected light is not received by the ranging sensor, or when a measurement value measured by the ranging sensor is longer than a predetermined value.

Abstract: A system and method of vibration and audible noise reduction in a LiDAR resonator includes a spring fork mechanism including multiple spring forks. Each spring fork includes two tines. The first tine of a first and second spring fork include a mounted optical module to transmit a light pulse and receive a reflection of the light pulse. The second tine of the first and second spring forks include a mounted counterweight having a mass and center of gravity equal to a mass and center of gravity of the mounted optical module. To reduce or eliminate longitudinal vibrations each tine includes a first section and a second section, the first section attached to the second section by a U-shaped section.

Abstract: Provided is a spatial light modulator configured to modulate light, the spatial light modulator including a first reflective layer, a resonance layer on the first reflective layer, and a second reflective layer on the resonance layer, the second reflective layer including a plurality of grating structures space apart from each other, wherein the plurality of grating structures include silicon (Si) having an extinction coefficient k that is less than or equal to 1e-5 with respect to light in the predetermined wavelength band.

Abstract: A LIDAR transmitter system comprising an array of laser energy sources, the laser energy sources being arranged on a first curved surface and being configured to emit laser energy towards a LIDAR target. The LIDAR transmitter system further comprising at least a first lens arranged in the optical path between the array of laser energy sources and the LIDAR target, wherein the first curved surface is positioned at an image plane of the first lens.

Abstract: A sensor device comprising a retro-reflecting element having a front boundary surface configured for entry of measuring light into the retro-reflecting element and a back boundary surface configured for reflecting a first part of the measuring light as reflected measuring light and configured for transmitting a second part of the measuring light as transmitted measuring light, wherein the front boundary surface and the back boundary surface are arranged on opposite sides of the retro-reflecting element, and a sensor unit configured and arranged so that the transmitted measuring light is detectable by the sensor unit. The front boundary surface and the back boundary surface are of curved shape.

Abstract: A machine-readable symbol reader can capture a cubic data set of an object carrying a machine-readable symbol. The machine-readable symbol reader can include a hyperspectral sensor, and the cubic data set may include image data for the machine-readable symbol within multiple, different wavelengths of the electromagnetic spectrum. The cubic data may be separated into portions that are each analyzed by specific processors, and those separate analyses can be used to increase the capabilities and efficiency compared to known machine-readable symbol readers.

Abstract: Disclosed herein are optical systems (e.g., LiDAR systems) and methods with improved eye safety. In some embodiments, a system includes a first light emitter configured to illuminate a first field of view (FOV) using light emitted at a first wavelength and a second light emitter configured to illuminate a second FOV using light emitted at a second wavelength. The second FOV is wider than the first FOV, and the first FOV extends to a further distance from the system than the second FOV. The system also includes a sensor configured to detect reflections off of targets within the second FOV, and at least one processor configured to execute one or more machine executable instructions.

Abstract: A light detection and ranging (LIDAR) signal processing apparatus capable of accurately measuring a distance to an object using an optical means may include: a flight length deriving unit to calculate a flight length of a first laser light emitted from a sensor toward a road with an inclination, a vertical inclination measurement unit to estimate a vertical inclination of the vehicle with respect to the road based on the flight length, a horizontal inclination measurement unit to estimate a horizontal inclination of the vehicle with respect to the road based on the flight length, a reliability determination unit to determine reliability of information detected by the sensor based on the vertical inclination and the horizontal inclination of the vehicle, and a data processing unit to process the detected information when the detected information is reliable.

Abstract: A method for manufacturing a LiDAR device is proposed. The method may include providing a LiDAR module including a laser emitting module and a laser detecting module to a target region. The method may also include adjusting, on the basis of first detecting data obtained from the laser detecting module, a relative position of a detecting optic module with respect to the laser detecting module. The method may further include adjusting, on the basis of image data obtained from at least one image sensor, a relative position of an emitting optic module with respect to the laser emitting module.

Abstract: A wireless system capable of accurately and inexpensively measuring a distance between a vehicle and a mobile terminal is provided. When a first sum of a first received signal strength indication and a second received signal strength indication is larger than a second sum of a third received signal strength indication and a fourth received signal strength indication, a distance between the vehicle and the mobile terminal is obtained from a phase difference between a first radio wave and a second radio wave. When the first sum of the first received signal strength indication and the second received signal strength indication is smaller than the second sum of the third received signal strength indication and the fourth received signal strength indication, the distance between the vehicle and the mobile terminal is obtained from a phase difference between a third radio wave and a fourth radio wave.

Abstract: A control device for a ceiling fan may have a motor drive circuit configured to control a rotational speed of a motor of the ceiling fan, an occupancy sensing circuit, and a control circuit configured to adjust the rotational speed of the motor in response to a detected occupancy or vacancy condition. The control circuit may process the signals generated by the occupancy sensing circuit to eliminate the effects of vibrations and/or wobbling of the ceiling fan. The control circuit may control the motor drive circuit to adjust the rotational speed of the motor in response to an accelerometer to minimize the magnitude of the wobble of the ceiling fan. The control circuit may be configured to learn a preferred rotational speed for the motor. The control circuit may also be configured to control the rotational speed of the motor to affect a thermal comfort level of an occupant.

Abstract: The present disclosure relates to a pulse radar device comprising a memory that includes a first memory and a second memory, each storing a scan vector. The device also includes a clock generator that produces a transmission clock signal and a reception clock signal. The reception clock signal is generated by delaying the transmission clock signal by a clock delay value. Furthermore, the device includes a transmitter that generates a transmission pulse by accepting the transmission clock signal from the clock generator and emits the pulse. Additionally, the device includes a receiver that receives an echo pulse reflected from a target by accepting the reception clock signal from the clock generator. The receiver then calculates the received echo pulse to generate a representative scan vector.

Abstract: A position estimation method is performed by a position estimation apparatus, wherein the position estimation method includes switching from a UWB static domain to a UWB dynamic domain when a startup state of a vehicle is switched from OFF to ON; operating one or more anchors mounted on the vehicle; and estimating a position of a tag by performing at least one of an advanced trilateration measurement (ATM) algorithm and a smart anchor positioning (SAP) algorithm, wherein the one or more anchors and the tag transmit and receive ultra wideband (UWB) pulse waves.

Abstract: A vehicular radar sensing system includes a radar sensor that includes a radiation slot sheet, a wave port sheet, and a waveguide sheet sandwiched between the radiation slot sheet and the wave port sheet. The sheets are joined together. The radiation slot sheet includes at least one transmitter that transmits radio signals and at least one receiver that receives radio signals. The waveguide sheet guides the transmitted radio signals from the radiation slot sheet to the wave port sheet and guides reflected radio signals from the wave port sheet to the radiation slot sheet. The transmitted radio signals are provided to the environment via at least one wave port of the wave point sheet. The reflected radio signals are received from the environment at the at least one wave port.

Abstract: A radar control device may comprise a transceiver configured to transmit a transmission signal for detecting surroundings of a host vehicle and receive a reception signal reflected back to the transceiver, a track generator configured to generate a track for an object detected based on the reception signal, a vehicle position calculator configured to calculate a position of the host vehicle at every preset period, and an orientation angle estimator configured to estimate an orientation angle of the track based on the track for the detected object and the position of the host vehicle.

Abstract: Methods, apparatus and systems for wireless sensing are described. In one example, a described system in a wireless data communication network comprises: a transmitter configured to transmit a time series of at least one wireless sounding signal (WSS) based on a standard wireless network protocol associated with the wireless data communication network, and a receiver. The wireless data communication network comprises a physical (PHY) layer, a medium access control (MAC) layer, and at least one higher layer. The receiver is configured to: receive the time series of at least one WSS (TSWSS) based on the standard wireless network protocol through a wireless channel of a venue, and perform a plurality of wireless sensing measurements based on the received TSWSS to obtain sensing measurement results based on the standard wireless network protocol.

Abstract: An indoor positioning method includes scanning for registered Wi-Fi nodes with known coordinates to generate a list of the registered Wi-Fi nodes. The method also includes performing a ranging operation by (i) selecting nodes to range with from the list of the registered Wi-Fi nodes, and (ii) processing ranging responses from the selected nodes to generate a series of distance measurements. The method further includes obtaining a series of sensor readings generated by one or more inertial measurement units (IMUs) of a device. The method also includes estimating a position of the device based on the series of distance measurements and the series of sensor readings using first and second filtering operations that are performed in parallel.

Abstract: A ranging method includes: receiving, by a ranging user equipment (UE), a ranging service request sent by a ranging initiating UE, wherein the ranging service request includes an identifier and a ranging parameter; determining, by the ranging UE, a ranging role of the ranging UE according to the identifier, wherein the ranging role includes an observation UE or a target UE; and performing, by the ranging UE, ranging according to the ranging parameter and the ranging role.

Abstract: Joint communication and sensing by a joint communication and sensing system in a wireless network is disclosed. A transmitter is arranged to transmit a first beam in a direction selected from a plurality of directions stored in a memory, where each direction corresponds to a direction of a respective remote device. The first beam comprises communication symbols to be communicated to the remote device in the direction during a communication session with the remote device. A reflection of the transmitted first beam is received via a receive antenna during the communication session, where the reflected first beam comprises the communication symbols. A position of one or more objects is identified based on a timing of transmission and receipt of the communication symbols in the transmitted first beam and the reflected first beam respectively and the sense symbols in the transmitted second beam and the reflected second beam respectively.

Abstract: Disclosed are systems and methods for scintillation-based neural network for RADAR target classification. In some aspects, a method includes generating a point cloud from radio frequency (RF) scene responses received from a radio detection and ranging (RADAR) sensor for a scanned scene; populating a rolling buffer with frame data from the point cloud, the frame data including radar cross section (RCS) values, RCS scintillation measurements, and velocity values for objects in the scanned scene; inputting the RCS scintillation measurements and velocity values for an object of the objects to a convolutional neural network (CNN); and receiving a classification of the object from the CNN, wherein the CNN is to utilize a probability density function (PDF) estimate of the RCS scintillation measurements and the velocity values to determine fits with one or more reference PDFs based on a Neyman-Pearson evaluation, and wherein the fits are assessed to classify the object.

Abstract: Disclosed are a process (1) and a device (2) for generating at least one synthetic radio frequency, RF, image for machine learning. The process (1) comprises: having (11) a three-dimensional, 3D, body model of a human; sampling (12) the 3D body model; and generating (13) the at least one synthetic RF image in accordance with an imaging transformation of the 3D body sample. This provides labeled data in the form of RF images for training of machine learning algorithms.

Abstract: Provided are methods for generation of representations of radar data. Some methods described include: receiving ADC raw data of a radar sensor of a vehicle; performing range FFT, Doppler FFT, and azimuth FFT on the ADC raw data; generating a 1D range heat map tensor representing the range FFT, a 2D RD heat map tensor representing a combination of the range FFT and the Doppler FFT, a 2D RA heat map tensor representing a combination of the range FFT and the azimuth FFT, or a 3D RAD matrix tensor representing a combination of the range FFT, the Doppler FFT, and the azimuth FFT; and inputting at least one of the 1D range heat map tensor, the 2D RD heat map tensor, the 2D RA heat map tensor, or the 3D RAD matrix tensor, to a machine learning model for detecting objects on a road network around the vehicle.

Abstract: Systems and methods are provided for synchronizing a bistatic inverse synthetic aperture radar (ISAR) imaging radar system using only the signal that travels directly between the transmitter and receiver. In an embodiment, an extracted direct path signal is used as a matched filter reference function. In an embodiment, full pulse compression is used for a matched filter implementation. Embodiments of the present disclosure provide systems and methods for detecting and extracting the direct path signal and using this direct path signal to synchronize the received signal stream in a bistatic ISAR system.

Abstract: A method and device for object recognition for information collected from a sensor of a moving object are disclosed. The method may include identifying a target with a confidence level of recognition accuracy less than a threshold, based on first sensing information collected from a moving object, obtaining second sensing information on the target from an interworking moving object, synchronizing the first sensing information and the second sensing information based on time information included in the first sensing information and time information included in the second sensing information, and training a neural network for estimating recognition information of the target from the first synchronized sensing information by using information on the target included in the second synchronized sensing information as ground truth data.

Abstract: The present subject matter relates to a radar system for the detection of surroundings, in particular for a vehicle and/or a transport device, and/or for stationary application, comprising: at least one transmitting-receiving unit for transmitting and receiving radar signals, which is configured to transmit a plurality M of physical angle-modulated signals, in particular chirps, from which a plurality N of virtual angle-modulated signals, in particular chirps, can be formed, wherein each virtual signal comprises several, in particular M, sampling points which are distributed over the physical chirps.

Abstract: A method of predicting a movement of a cut-in object, including separating a Doppler velocity corresponding to at least one wheel of the cut-in object from radar data based on a velocity of the cut-in object, determining a position of the at least one wheel of the cut-in object based on the Doppler velocity, generating at least one of first movement information related to a horizontal movement of the cut-in object or second movement information related to a moving direction of the cut-in object based on the position of the at least one wheel, and determining a position of the cut-in object based on at least one of the first movement information or the second movement information.

Abstract: A method for providing training datasets for training an object classification model for object classification in an ultrasonic sensor system is disclosed.

Abstract: Systems and methods for network-based ultrasound imaging are provided, which can include a number of features. In some embodiments, an ultrasound imaging system images an object with three-dimensional unfocused pings and obtains digital sample sets from a plurality of receiver elements. A sub-set of the digital sample sets can be electronically transferred to a remote server, where the sub-set can be beamformed to produce a series of two-dimensional image frames. A video stream made up of the series of two-dimensional images frames can then be transferred from the remote server to a display device.

Abstract: A transducer mounting device for positioning and orienting a transducer is disclosed that can pivot the transducer about multiple axes and readily provide information regarding a direction which the transducer is pointing when the transducer detects a submerged mass. The disclosed device includes a rotatable pole to which a microcontroller, transmission module, and articulating sensor module are attached. Each of the transmission module and articulating sensor module include motors for positioning and orienting the transducer.

Abstract: A target object detection apparatus (20) includes an acquisition unit (210), a likelihood information generation unit (220), and a determination unit (230). The acquisition unit (210) acquires an IF signal from an irradiation apparatus (10) at a plurality of timings. The likelihood information generation unit (220) executes, by processing the IF signal, processing of generating, with respect to each of a plurality of IF signals, likelihood information for a region (hereinafter, referred to as a target object region) having a possibility in that an accompaniment, i.e., a target object is present. The likelihood information indicates a distribution of presence probabilities of a target object in at least a height direction in a target object region. Then, the determination unit (230) determines a presence or absence of a target object, by using likelihood information generated for each of a plurality of target object regions.

Abstract: A method is for detecting one or more objects in a detection zone using a time-of-flight sensor. The method includes emitting optical radiation via the emission circuitry of the sensor and subsequently capturing the reflected optical radiation using the reception circuitry. This captured radiation is quantified in terms of photons, and measurement circuitry determines both the amount of these photons and the distance from the sensor to the object(s). An analysis of the photon count, combined with the calculated distance, is used to determine the presence or absence of objects within the detection zone.

Abstract: A light receiving device (20) according to one aspect of the present disclosure includes: a light receiving section (22) including a plurality of photon-counting light receiving elements that receives reflected light from a distance measurement target (40) based on irradiation pulsed light from a light source section (10); a selecting section (23) that selects individual detection values of the plurality of light receiving elements at a predetermined time; an addition section (24) that generates 2N?1 binary values (N is a positive integer) from the individual detection values of the plurality of light receiving elements at the predetermined time selected by the selecting section (23) and that calculates an N-bit pixel value by adding up all the 2N?1 binary values; and a computing section (26) that performs computation related to distance measurement using the N-bit pixel value calculated by the addition section (24).

Abstract: A distance measurement device determines, out of pixels, first pixels that are readout targets and second pixels that are not readout targets; performs, on each of the first pixels, exposure operations of n types at different timings where n is an integer of at least 4; and reading out signal charges of n types to n packets in each of vertical transfer portions, the signal charges of the n types being generated through the exposure operations of the n types. Within a region in which the vertical transfer portions are disposed, a first region in which the n packets are created is adjacent to a pixel region in which m pixels are consecutively disposed in a vertical direction where m is an integer of at least 3. The m pixels include at least one of the first pixels and at least one of the second pixels.

Abstract: A ranging device includes: a pulse generation unit configured to generate a signal including a pulse based on incident light; a first decoder unit configured to generate a first frequency distribution having a first class width based on the time count value and the number of pulses; a second decoder unit configured to generate a second frequency distribution having a second class width narrower than the first class width based on the time count value and the number of pulses; a first ambient light information generation unit configured to generate first ambient light information based on the first frequency distribution; and a distance calculation unit configured to calculate distance information based on the second frequency distribution.

Abstract: An annotation system uses annotations for a first set of sensor measurements from a first sensor to identify annotations for a second set of sensor measurements from a second sensor. The annotation system identifies reference annotations in the first set of sensor measurements that indicates a location of a characteristic object in the two-dimensional space. The annotation system determines a spatial region in the three-dimensional space of the second set of sensor measurements that corresponds to a portion of the scene represented in the annotation of the first set of sensor measurements. The annotation system determines annotations within the spatial region of the second set of sensor measurements that indicates a location of the characteristic object in the three-dimensional space.