Fujitsu Patente

An information processing apparatus acquires measurement data indicating measurement values of quantum states of qubits measured after a quantum computer has repeatedly executed, a plurality of times, a quantum gate sequence including a plurality of quantum gates for the qubits. The information processing apparatus divides a space to which a matrix indicating a quantum operation by the quantum gate sequence belongs into a first subspace including a singular point at which eigenvalues satisfying predetermined conditions are calculated, and a second subspace that does not include the singular point, and generates, within a range of the second subspace, a function that linearly approximates an influence of an error of a quantum gate included in the plurality of quantum gates on the measurement value. The information processing apparatus estimates, by using the function and the measurement data, the error of the quantum gate when the quantum computer executes the quantum gate.

A processing unit groups a plurality of sequences to be aligned in a multiple sequence alignment problem into a plurality of groups, selects a representative sequence from each of the plurality of groups, and causes an Ising machine to perform a first alignment process among the representative sequences of the plurality of groups. The processing unit replaces the representative sequence before the first alignment process with the corresponding representative sequence obtained through the first alignment process in each of the plurality of groups, and causes the Ising machine to perform a second alignment process within each of the plurality of groups after the replacement of the representative sequences.

A quantum device (100) includes a quantum chip (10) including a substrate (11) having a first surface (12) and a second surface (13), a quantum bit (15) provided on the first surface, and an electrode (18) electrically connected to the quantum bit and provided on the second surface, a mounting portion (30) having a mounting surface (32) on which a peripheral edge portion of the quantum chip is mounted, the mounting portion having a linear expansion coefficient different from that of the substrate, a conductor pin (40) having a tip end in contact with the electrode, and a holding portion (50) fixed to the mounting portion, the holding portion holding the conductor pin and having a linear expansion coefficient different from that of the substrate, wherein the electrode has a shape having a longitudinal direction in a direction radially extending from a center of the substrate in plan view, and the quantum chip is mounted on the mounting surface by inserting a protrusion (31) provided on one of the substrate and the mounting surface into a hole (19) provided on another of the substrate and the mounting surface, the hole having a longitudinal direction in a direction extending radially in plan view.

An information processing program for causing a computer to execute a process, the process includes: obtaining a decomposition number of a Trotter decomposition approximating a quantum unitary corresponding to an Ising model; and setting a quantum circuit expressing a formula of the Trotter decomposition by one or more first partial circuits and a second partial circuit, the one or more first partial circuits being of a first count one less than the obtained decomposition number and each including an Rz gate that represents a rotation action about a Z-axis on a qubit, the second partial circuit being free of the Rz gate and coupled to a rear of the one or more first partial circuits.

An information processing program causes a computer to execute a process including detecting a bounding box of an object in a plurality of frames captured by a camera installed at a predetermined point, and generating a heat map that represents average reliability at each set of coordinates in a range captured by the camera, based on reliability of each of a plurality of the detected bounding boxes.

A processing unit acquires the data of first and second classical shadows corresponding to first and second quantum data represented by qubits, and geometric structure information indicating the geometric structure of a space including lattice points associated with the qubits. The first and second classical shadows include first and second data elements, respectively, corresponding to the qubits. The processing unit sets local subspaces each including a set of local lattice points in the space, based on the geometric structure information. The processing unit calculates, for each local subspace, a value of a first kernel function based on first and second data elements of the first and second classical shadows corresponding to the lattice points of that local subspace, and calculates a value of a second kernel function corresponding to the first and second classical shadows, based on the values of the first kernel function corresponding to the local subspaces.

A quantum device includes a first qubit circuit, a second qubit circuit, and a cover chip. The first qubit circuit includes a first port, a first qubit, and a first resonator coupled to the first qubit and the first port and having a first resonant frequency. The second qubit circuit includes a second port, a second qubit, and a second resonator coupled to the second qubit and the second port and having a second resonant frequency different from the first resonant frequency. The cover chip includes a waveguide coupled to the first port and the second port.

An information processing program for causing a computer to execute a process including: setting first and second functions, for a quantum circuit used when solving a combinatorial optimization problem and including, for each layer, a first partial circuit representing an action of a mixer unitary operator and a second partial circuit representing an action of a cost unitary operator, the first function representing a first variational parameter of the first partial circuit, the second function representing a second variational parameter of the second partial circuit; and calculating a solution to the combinatorial optimization problem by updating a value of a first transformation parameter related to the first function and a value of a second transformation parameter related to the second function so as to optimize an expected value of a cost function corresponding to the combinatorial optimization problem by using the quantum circuit after setting the first function and the second function.

A method of manufacturing a Josephson junction element includes forming a lower electrode (20) on a substrate (10) by using a superconducting material, depositing a first insulating film (30) on the substrate to cover the lower electrode, depositing a superconducting film on the first insulating film, and forming an upper electrode (40) by etching the superconducting film, the upper electrode covering an entire lower electrode in plan view and having an outer peripheral edge located outside an outer peripheral edge of the lower electrode.

A method of manufacturing a quantum device (100, 200) includes bonding a diamond substrate (30) including a color center (21) to a layer (32) provided on a support substrate (10), forming a first portion (40) including the color center, and a second portion (42) having an inclined surface (43) inclined with respect to a side wall (41) of the first portion by etching the diamond substrate after the bonding the diamond substrate, forming a metal film (45) on the inclined surface, and forming a first optical waveguide portion (20) by cutting the first portion at a first position farther from the layer than the color center by emitting a laser beam reflected by the metal film onto the side wall of the first portion.

A quantum device (1) includes a qubit substrate (170) having qubits and conductive pads coupled to the qubits, the qubit substrate (170) having a first thermal expansion coefficient; a first support (210) having a second thermal expansion coefficient; a plurality of second supports (220) disposed between the qubit substrate (170) and the first support (210), each having a third thermal expansion coefficient a plurality of coupling members (230) each coupling the first support (210) and a corresponding second support of the second supports (220); and spring pins (310) supported by the second supports and contacting the conductive pads. A control signal for controlling a state of each of the qubits is supplied to a corresponding conductive pad of the conductive pads through a corresponding spring pin of the spring pins (310). An absolute value of a difference between the first thermal expansion coefficient and the second thermal expansion coefficient is smaller than an absolute value of a difference between the first thermal expansion coefficient and the third thermal expansion coefficient.

A generation program causes a computer to execute a process including acquiring purchase information of a product or service, detecting an area in a store where a product is arranged, when an analysis result of the acquired purchase information during a certain period of time set in advance satisfies a predetermined condition, among a plurality of video images of the detected area, extracting an analysis result of each of the video images according to the certain period of time, every time the analysis result of the video image according to the certain period of time is extracted, storing the extracted analysis result of the video image in a storage, by inputting a prompt including the analysis result of the video image stored in the storage during a first period of time into a language model, and generating and outputting information on a measure to be taken in the store.

An information notification program causes a computer to execute a process including acquiring domain knowledge information including a work procedure of an operation to be monitored, analyzing video to be monitored and first identifying an image region of a person and an image region of an object to be worked by the person in the video, inputting a prompt composed of a feature of the identified image region of the person, a feature of the identified image region of the object, and the domain knowledge information to a large multi-modal model, and causing, when a work performed by the person in the video is not a behavior based on the work procedure of the operation, the large multi-modal model to generate information related to the work procedure, and notifying the generated information related to the work procedure.

A provider terminal apparatus encrypts a vector obtained by vectorizing a plurality of pieces of provided date with homomorphic encryption to generate indexes, generates a plurality of pieces of encrypted data obtained by encrypting the pieces of provided data, and transmits a search table in which the indexes and the encrypted data are included to an information search server (1). A searcher terminal apparatus transmits a search query obtained by encrypting a vector obtained by vectorizing query information with homomorphic encryption to the information search server (1). The information search server (1) performs homomorphic operation according to the homomorphic encryption between the search query and each of the indexes registered in the search table, creates a similarity list, and transmits the similarity list to the searcher terminal apparatus. The searcher terminal apparatus extracts a similar index similar to the search query based on the similarity list.

A computer-implemented method comprising: generating, based on input network packet data, a target flow summary comprising information relating to packet flow between two addresses during a target time period; generating, based on the target flow summary, a target latent space representation; identifying k reference latent space representations most similar to the target latent space representation, wherein the plurality of reference latent space representations correspond to a plurality of reference flow summaries; generating, using a first LLM and based on the target flow summary, a target flow report in natural language format and generating, using the first LLM and based on k reference flow summaries, k reference flow reports in natural language format; generating, using a second LLM and based on the target flow report and the k reference flow reports with their corresponding classifications, a classification of the packet flow as malicious or benign.

A specifying program causes a computer (200) to execute a process including acquiring an attribute of training data acquiring a document related to a decisionmaking condition for a specific attribute of a user inputting a prompt including the attribute of the training data and a token of the document related to the decisionmaking condition to a large-scale language model and outputting an appearance probability of a token related to the attribute and specifying a label of a training target of a machine learning model from the training data based on the output appearance probability of the token.

A method may include discretizing a partial differential equation with an initial value condition to obtain a linear equation. The method may also include scaling at least one discretization parameter of the linear equation to obtain a scaled linear equation. The method may include applying a quantum algorithm to the scaled linear equation to obtain a linear system with a preconditioned matrix. The method may further include performing a quantum singular value transformation to apply an inverse of the preconditioned matrix to obtain a solution. The method may include extracting an integral probability from the solution using an integral interpolation.

An information processing apparatus generates a second quantum circuit by converting a first two-qubit gate included in a first quantum circuit into a two-qubit gate sequence in which the gate operation of the first two-qubit gate is repeated a predetermined number of times consecutively. The information processing apparatus converts a second two-qubit gate, which forms a part of the two-qubit gate sequence included in the second quantum circuit, into a first equivalent circuit representing a gate operation with a first global phase. In addition, the information processing apparatus converts a third two-qubit gate, which forms a part of the two-qubit gate sequence, into a second equivalent circuit representing a gate operation with a second global phase. Thus, a third quantum circuit is generated. The information processing apparatus then causes a quantum computer to execute the third quantum circuit.

A search unit performs a search process. In the search process, a first process and a second process are iteratively performed. The first process is a process of selecting a change-target state variable from a plurality of state variables included in an evaluation function, based on the amounts of change in the value of the evaluation function caused by updating the values of the individual state variables and a first function that is a function representing a selection criterion for the change-target state variable according to the amounts of change. The second process is a process of updating the value of the selected change-target state variable. A control unit switches the function used by the search unit in the search process to a second function different from the first function.

In executing trials each of which changes the value of a state variable included in state variable groups included in an evaluation function for a combinatorial optimization problem, a processing unit performs, based on evaluation function information on the evaluation function, a first process of determining a candidate for a first state variable whose value is to be changed in parallel for each state variable group, in each trial, with a first probability, performs, based on the evaluation function information, a second process of selecting one second state variable and determining whether to accept a change in the value thereof in parallel for each state variable group, in each trial, with a second probability, updates the value of the first state variable selected from the candidates determined by the first process, and updates the value of a second state variable whose change is accepted by the second process.

A detection program causes a computer to execute a process including converting input tabular data into image data, generating a first machine learning model by machine learning using the tabular data as training data, and generating a second machine learning model by machine learning using the image data as training data, and detecting presence of an adversarial attack on input data based on a first prediction result for the input data obtained using the first machine learning model and a second prediction result for the input data obtained using the second machine learning model.

A training program causes a computer to execute a process including: first inputting a first image capturing a target compound to an encoder of an auto-encoder including a latent space that is isometric with respect to an input space; second inputting a latent variable output by the encoder and a typical compound model corresponding to a typical case of a three-dimensional structure of the target compound to a decoder of the auto-encoder; and updating parameters of the encoder and the decoder, based on a reconfiguration error between a second image reconfigured based on an output of the encoder and the first image.

An information processing program for causing a computer to execute a process including: generating a plurality of first quantum circuits by a local compilation method, the first quantum circuits representing an action of a time evolution operator for a first time period and having a depth smaller than a depth of a first target quantum circuit representing the action of the time evolution operator for the first time period; and generating a second quantum circuit by the local compilation method, the second quantum circuit being smaller in depth than a second target quantum circuit obtained by combining two or more of the generated first quantum circuits and representing the action of the time evolution operator for a second time period longer than the first time period, the second quantum circuit representing the action of the time evolution operator for the second time period.

An information processing apparatus expands an imaginary time evolution equation for calculating a thermal-equilibrium expectation value into equations of a plurality of degrees, generates sets of the degrees, and generates, for each set, a quantum circuit indicating quantum computation of a physical quantity value obtained by partial imaginary time evolution. A quantum computer is caused to repeatedly execute quantum computation based on the quantum circuit for each set. When doing so, a state in a stationary distribution of measurement results obtained by quantum computation based on a second quantum circuit corresponding to a second set is used as an input state at the start of quantum computation based on a first quantum circuit corresponding to a first set. The thermal-equilibrium expectation value of the physical quantity at the finite temperature is then calculated based on the values of the physical quantity after convergence for each set.

A reinforcement learning program that causes a computer to execute a process including: executing reinforcement learning of a machine learning model (21a), using a first three-dimensional (3D) molecular structure of a first compound to be adapted to a first object as a state, assignment of a type of an additional atom, a binding position of the additional atom to the first 3D molecular structure, and a 3D position of the additional atom, to the first compound in the state as an action, and an index to evaluate a second 3D molecular structure obtained by binding the additional atom to the first 3D molecular structure in the action as a reward.

A display program causes a computer (1) to execute a process including based on a first dataset, generating (S13) a second dataset that is virtual and has a causal relationship between variables included in the first dataset, in accordance with an estimation result indicating the causal relationship, calculating (S14) reliability of each inter-variable relationship between a first variable serving as a cause in the causal relationship and a second variable serving as an effect in the causal relationship, based on difference between the first dataset and the second dataset and displaying (S16) the calculated reliability of each inter-variable relationship.

An information processing program for causing a computer to execute a process, the process including: classifying a plurality of items in to a plurality of groups; searching for a causal order so as to increase a likelihood of an estimation result of a regression model, the causal order indicating a presence of a causal relationship in a direction from one group to another group between at least some of the plurality of groups, the regression model estimating a value of a downstream item based on a value of an upstream item in a causal relationship; and outputting a directed acyclic graph generated based on the discovered causal order and representing a presence of a causal relationship in a direction from one item to another item, between at least some of the plurality of items.

An information processing program causes a computer to execute a process, the process including generating a causal graph coupling a plurality of nodes by directed edges based on a plurality of data serving as a basis in generating the causal graph, the causal graph being generated so as to optimize a value of an objective function that includes a first term indicating that the more destination connections and connection sources shared between different nodes in the causal graph, a higher is an evaluation thereof; and updating the generated causal graph so as to aggregate, among the plurality of nodes, two or more nodes sharing a common connection destination and a common connection source, the two or more nodes being aggregated into a single node.

The present disclosure relates to a method, a device, and a storage medium for multiple object tracking. According to an embodiment of the present disclosure, the method comprises: determining a plurality of object whole-body detection boxes in a current input image by performing object whole-body detection, and, determining a plurality of object head detection boxes in the current input image by performing object head detection; determining whole-body identifiers of the plurality of object whole-body detection boxes by performing whole-body trajectory association; determining head identifiers of the plurality of object head detection boxes by performing head trajectory association; determining a plurality of object whole-body prediction boxes corresponding to the plurality of object head detection boxes based on positions and sizes of the plurality of object head detection boxes; determining object whole-body association boxes of the plurality of object head detection boxes in the plurality of object whole-body detection boxes based on areas occupied by the plurality of object whole-body prediction boxes corresponding to the plurality of object head detection boxes in the input image; and updating the object whole-body trajectory set based on whole-body identifiers of the object whole-body association boxes.

An information processing device having a controller configured to: when iterations of updating parameters defining a variational quantum circuit are executed according to a variational quantum eigensolver method, determine for each first parameter, a number of level values settable as a value thereof based on a first degree of contribution of the first parameter to an amount of change in energy corresponding to a quantum state represented by the variational quantum circuit; calculate for each first parameter, a second degree of contribution thereof to the variational quantum eigensolver method, based on the energy in each of two or more patterns representing combinations of level values to be set as the value of the first parameter when setting one of the level values of the number determined for the first parameter; and set, as parameters to be updated, each second parameter that among the first parameters has a relatively high second degree of contribution.

An active space determination program causes a computer to execute a process including: acquiring occupation number data including data indicating time-series change of an electron occupation number in each of a plurality of molecular orbitals; computing a solution of a mathematical optimization model formulated with a constraint and an objective function, using the occupation number data as input, the constraint being that a sum of electron occupation numbers in the molecular orbitals is constant, the objective function giving a high evaluation to selection of a combination of molecular orbitals in which the time-series change of the electron occupation number has a negative correlation, relative to a combination of molecular orbitals in which the time-series change of the electron occupation number has a positive correlation, when selecting a subset to be used in quantum chemical calculation from among the molecular orbitals; and outputting a result of the computation.

A detection program causes a computer to execute a process includes acquiring an input image from a camera acquiring a first detection result by inputting the input image to a first detection model that performs object detection acquiring, a plurality of detection results by inputting the input image to a plurality of detection models trained based on different training data sets with a parameter of the first detection model used as an initial value, and generating a detection result in which the first detection result and the detection results are combined.

A processing unit selects a first constraint and performs a search process repeating first and second processes. The first process determines, based on a first local field, whether to allow a change in a first state variable value. If allowed, the second process updates the first local field, based on a first weight coefficient for the first state variable, updates a second local field for the first constraint, based on a second weight coefficient for the first state variable and the first constraint, and updates the first local field, based on the second local field before and after updating. The processing unit updates a second local field for a second constraint unused, based on changes in state variable values caused by the search process and a second weight coefficient for the second constraint, and corrects the first local field, based on the second local field before and after updating.

A quantum computing apparatus (10) includes a plurality of substrates (20, 30, 40) that are stacked and include a substrate (20) on which a qubit device (11) is provided, and through-vias (35A, 35B, 45A, 45B) provided in at least a substrate (40) disposed in an uppermost layer and a substrate (30) disposed in a lowermost layer, among the plurality of substrates (20, 30, 40), and electrically coupled to the substrate (20) disposed in an adjacent layer.

A processing unit determines a first solution to a surrogate problem represented by a second evaluation function obtained by adding a product of an auxiliary variable and a weight coefficient corresponding to a constraint to a first evaluation function of a mixed-integer programming problem including integer and continuous variables, using the branch and bound method with linear relaxation of the integer variable. The processing unit determines a second solution to the mixed-integer programming problem by local search using an initial solution where the integer variable in the first solution is set to an integer value while fixing the continuous variable value, determines a third solution to the mixed-integer programming problem while fixing the integer variable value included in the second solution, decreases the weight coefficient corresponding to the constraint satisfied by the third solution, and then repeats determining the first, second, and third solutions and decreasing the weight coefficient.

An information processing apparatus includes a processor configured to generate a set of a plurality of routes for a plurality of moving bodies that depart from a specific node and return to the specific node via a plurality of nodes, the plurality of routes representing an order in which the plurality of moving bodies visit the nodes generate, by using the generated set of routes, a route selection problem with constraints on a shipment volume of a node serving as a shipment source of packages solve the generated route selection problem and output a result of solving the route selection problem.

A computer implemented method for detecting computer generated images comprising: loading an input image, inputting the input image and a representation describing the input image into a denoising model for denoising the input image using the representation, generating a denoised image embedding from the denoised image, generating an input image embedding from the input image, comparing a difference between the input image embedding and the denoised image embedding with a similarity decision threshold to determine whether the input image is a real image or a computer generated image.

A first wiring made of a superconductor and a first electrode made of a superconductor connected to the first wiring are formed on a substrate. A resist mask that covers at least the first wiring is formed. A surface of the first electrode using the resist mask is etched. An insulator film on the surface of the first electrode using the resist mask is formed. A second electrode made of a superconductor on the surface of the insulator film using the resist mask is formed.

A computer-implemented method for traffic simulation, the method comprising: executing a preliminary agent-based simulation model to generate a preliminary simulation prediction of traffic in a geographic area; accepting input of traffic data for the geographic area; performing data assimilation on the preliminary simulation prediction and the traffic data to generate assimilated traffic data.

A computer-implemented method comprising: generating, based on a plurality of candidate output tokens generated by a large language model, LLM, a plurality of candidate token embeddings, respectively and generating, based on at least one string describing prohibited content, at least one content embedding; comparing the candidate token embeddings with the at least one content embedding to determine for each candidate output token a similarity score indicating similarity between the candidate output token and the prohibited/unwanted content; for each candidate output token, generating, based on the similarity score and based on a selection probability assigned by the LLM to the candidate output token, an output score; and selecting, based on the output scores, a candidate output token for output by the LLM.