# US 12,002,265 B2 · Characterizing liquid reflective surfaces in 3D liquid metal printing > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,002,265 B2 - Document ID: US12002265B2 - Status: Granted - Kind: B2 - Application number: 17/356,604 - Filed: 2021-06-24 - Published: 2022-12-29 - Granted: 2024-06-04 - Earliest claimed: 2021-06-24 - Assignees: Xerox Corporation - Applicants: Xerox Corporation - Inventors: Robert R. Price; Raja Bala; Svyatoslav Korneev; Christoforos Somarakis; Matthew Shreve; Adrian Lew; Palghat Ramesh - Web page: https://patentgrep.com/patent/US12002265B2 ## CPC - G06V 20/46: {Extracting features or characteristics from the video content, e.g. video fingerprints, representative shots or key frames} - B41J 2/04536: {using history data} - G06F 18/214: Generating training patterns; Bootstrap methods, e.g. bagging or boosting - G06F 18/24147: {Distances to closest patterns, e.g. nearest neighbour classification} - G06N 3/0455: Auto-encoder networks; Encoder-decoder networks - G06N 3/0464: Convolutional networks [CNN, ConvNet] - G06N 3/0475: Generative networks - G06N 3/08: Learning methods - G06N 3/0895: Weakly supervised learning, e.g. semi-supervised or self-supervised learning - G06N 3/09: Supervised learning - G06V 20/49: {Segmenting video sequences, i.e. computational techniques such as parsing or cutting the sequence, low-level clustering or determining units such as shots or scenes} - G06N 3/082: modifying the architecture, e.g. adding, deleting or silencing nodes or connections - G06N 3/084: Backpropagation, e.g. using gradient descent - G06N 5/01: Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound - G06N 20/20: Ensemble learning - G06V 10/82: using neural networks - G06V 2201/07: Target detection - G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING - B41J: TYPEWRITERS; SELECTIVE PRINTING MECHANISMS, i.e. MECHANISMS PRINTING OTHERWISE THAN FROM A FORME; CORRECTION OF TYPOGRAPHICAL ERRORS - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS ## Abstract A method includes defining a model for a liquid while the liquid is positioned at least partially within a nozzle of a printer. The method also includes synthesizing video frames of the liquid using the model to produce synthetic video frames. The method also includes generating a labeled dataset that includes the synthetic video frames and corresponding model values. The method also includes receiving real video frames of the liquid while the liquid is positioned at least partially within the nozzle of the printer. The method also includes generating an inverse mapping from the real video frames to predicted model values using the labeled dataset. The method also includes reconstructing the liquid in the real video frames based at least partially upon the predicted model values. ## Claims ### Claim 1 A method, comprising: defining a model for a liquid while the liquid is positioned at least partially within a nozzle of a printer; synthesizing video frames of the liquid using the model to produce synthetic video frames; generating a labeled dataset that comprises the synthetic video frames and corresponding model values; receiving real video frames of the liquid while the liquid is positioned at least partially within the nozzle of the printer; generating an inverse mapping from the real video frames to predicted model values using the labeled dataset; reconstructing the liquid in the real video frames based at least partially upon the predicted model values; and extracting one or more metrics from the reconstructed liquid, wherein the metrics comprise a carrier oscillation frequency, a pulse-to-pulse covariance, a waveform decay rate, or a combination thereof. ### Claim 2 (depends on 1) The method of claim 1, wherein the model comprises a parametric model. ### Claim 3 (depends on 1) The method of claim 1, wherein the model values comprise: a relative weight of a function that describes an oscillation of the liquid; a constant offset; a constant representing a steady state shape of the liquid at rest; or a combination thereof. ### Claim 4 (depends on 1) The method of claim 1, wherein the predicted model values comprise a subset of the model values. ### Claim 5 (depends on 1) The method of claim 1, wherein generating the inverse mapping comprises training an artificial neural network using the labeled dataset to predict the predicted model values in the real video frames. ### Claim 6 (depends on 1) The method of claim 1, wherein generating the inverse mapping comprises selecting an entry from the labeled dataset that is most similar, in mean-square error, to the real video frames using direct nearest-neighbor matching. ### Claim 7 (depends on 1) The method of claim 1, wherein the liquid comprises a meniscus, and wherein reconstructing the liquid comprises reconstructing a shape, a motion, or both of the meniscus. ### Claim 8 (depends on 1) The method of claim 1, further comprising adjusting one or more parameters of the printer based at least partially upon the one or more metrics to adjust the liquid positioned at least partially within the nozzle. ### Claim 9 (depends on 8) The method of claim 8, wherein the one or more parameters comprise an amplitude, a frequency, or both of a waveform from a power source that causes the liquid to be jetted through the nozzle. ### Claim 10 A visual-based method for performing diagnostics on a printer, the comprising: defining a parametric model for a surface of a liquid while the liquid is positioned at least partially within a nozzle of the printer, wherein the surface of the liquid comprises a meniscus; synthesizing video frames of the meniscus of the liquid with a graphics simulator using the parametric model to produce synthetic video frames; generating a labeled dataset that comprises the synthetic video frames and corresponding parametric model values; receiving real video frames of the liquid while the liquid is positioned at least partially within the nozzle of the printer, wherein the synthetic video frames simulate the real video frames; generating an inverse mapping from the real video frames to predicted parametric model values using the labeled dataset, wherein the predicted parametric model values are a subset of the parametric model parameter values; reconstructing a shape and a motion of the meniscus of the liquid in the real video frames based at least partially upon the predicted parametric model values; and extracting one or more metrics from the reconstructed meniscus, wherein the metrics comprise a carrier oscillation frequency, a pulse-to-pulse covariance, a waveform decay rate, or a combination thereof. ### Claim 11 (depends on 10) The method of claim 10, wherein the parametric model values comprise: a first relative weight of a first Bessel function that describes an oscillation of the liquid; a second relative weight of a second Bessel function that describes the oscillation of the liquid; a constant offset; a constant representing a steady state shape of the meniscus at rest; or a combination thereof. ### Claim 12 (depends on 10) The method of claim 10, wherein generating the inverse mapping comprises training an artificial neural network using the labeled dataset to predict the predicted parametric model values in the real video frames. ### Claim 13 (depends on 10) The method of claim 10, wherein generating the inverse mapping comprises selecting an entry from the labeled dataset that is most similar, in mean-square error, to the real video frames using direct nearest-neighbor matching. ### Claim 14 (depends on 10) The method of claim 10, further comprising adjusting one or more parameters of the printer based at least partially upon the one or more metrics to adjust the meniscus. ### Claim 15 A method for characterizing a behavior of a liquid while the liquid is positioned at least partially within a nozzle of a printer, the method comprising: defining a parametric model for a surface of the liquid while the liquid is positioned at least partially within the nozzle of the printer, wherein the surface of the liquid comprises a meniscus; synthesizing video frames of the meniscus of the liquid with a graphics simulator using the parametric model to produce synthetic video frames; generating a labeled dataset that comprises the synthetic video frames and corresponding parametric model values, wherein the parametric model values comprise: a first relative weight of a first Bessel function that describes an oscillation of the liquid; a second relative weight of a second Bessel function that describes the oscillation of the liquid; a constant offset; and a constant representing a steady state shape of the meniscus at rest; jetting the liquid through the nozzle; illuminating the liquid with a light source when the liquid is positioned at least partially within the nozzle; receiving real video frames of the liquid while the liquid is positioned at least partially within the nozzle, wherein the synthetic video frames simulate the real video frames; generating an inverse mapping from the real video frames to predicted parametric model values using the labeled dataset, wherein the predicted parametric model values are a subset of the parametric model parameter values, and wherein generating the inverse mapping comprises: training an artificial neural network using the labeled dataset to predict the predicted parametric model values in the real video frames; or selecting an entry from the labeled dataset that is most similar, in mean-square error, to the real video frames using direct nearest-neighbor matching; reconstructing a shape and a motion of the meniscus of the liquid in the real video frames based at least partially upon the predicted parametric model values; extracting one or more metrics from the reconstructed meniscus, wherein the metrics comprise a carrier oscillation frequency, a pulse-to-pulse covariance, a waveform decay rate, or a combination thereof; and adjusting one or more parameters of the printer based at least partially upon the one or more metrics to adjust the meniscus. ### Claim 16 (depends on 15) The method of claim 15, wherein illuminating the liquid with the light source comprises illuminating the liquid with a plurality of light sources that are oriented at different angles with respect to the liquid. ### Claim 17 (depends on 16) The method of claim 16, wherein the light sources each have a different color. ### Claim 18 (depends on 15) The method of claim 15, wherein: the first relative weight has a value between −0.5 and 0.5; the second relative weight has a value between −0.1 and 0.1; the constant offset has a value between −1 and 1; and the constant has a value between 0 and 0.2. ### Claim 19 (depends on 15) The method of claim 15, wherein the one or more parameters comprise an amplitude, a frequency, or both of a waveform from a power source that causes the liquid to be jetted through the nozzle.