Inverse Task Research Articles

An efficient identity-preserving and fast-converging hybrid generative adversarial network inversion framework

In this paper, we present a novel Hybrid Generative Adversarial Network (HGAN) inversion framework that enables facial images to be rapidly inverted while preserving identity and personality characteristics. Accurate inversion of facial images requires high precision in computer vision and is critical to the success of future facial manipulations (age progression, regression, accessory, and hair stylization). However, existing methods often fail to preserve the personality characteristics of the real image, negatively affecting the accuracy of manipulations. In this context, our key contribution lies in using a transformer-based strategy to initiate the generator, which effectively models spatial relationships for detailed image processing. This approach is innovative because it leverages transformer structures to enhance image inversion tasks. Additionally, we introduce a novel loss function to enhance convergence speed and reliability, ensuring high accuracy in identity and personality trait preservation. Experimental results show that our method achieves a reconstruction accuracy of 93% and improves inversion time by 86%. This advancement could significantly impact facial manipulation technologies, laying the foundation for a technological breakthrough with potential applications in secure digital authentication systems and personal data protection. Our method may have a significant impact on privacy and security in future studies, contributing to the development of secure digital authentication systems and enhancing the protection of personal data. Therefore, our work is crucial for advancing the field of facial image manipulation and ensuring the privacy and security of personal data.

A Multi-Baseline Forest Height Estimation Method Combining Analytic and Geometric Expression of the RVoG Model

As an important parameter of forest biomass, forest height is of great significance for the calculation of forest carbon stock and the study of the carbon cycle in large-scale regions. The main idea of the current forest height inversion methods using multi-baseline P-band polarimetric interferometric synthetic aperture radar (PolInSAR) data is to select the best baseline for forest height inversion. However, the approach of selecting the optimal baseline for forest height inversion results in the process of forest height inversion being unable to fully utilize the abundant observation data. In this paper, to solve the problem, we propose a multi-baseline forest height inversion method combining analytic and geometric expression of the random volume over ground (RVoG) model, which takes into account the advantages of the selection of the optimal observation baseline and the utilization of multi-baseline information. In this approach, for any related pixel, an optimal baseline is selected according to the geometric structure of the coherence region shape and the functional model for forest height inversion is established by the RVoG model’s analytic expression. In this way, the other baseline observations are transformed into a constraint condition according to the RVoG model’s geometric expression and are also involved in the forest height inversion. PolInSAR data were used to validate the proposed multi-baseline forest height inversion method. The results show that the accuracy of the forest height inversion with the algorithm proposed in this paper in a coniferous forest area and tropical rainforest area was improved by 17% and 39%, respectively. The method proposed in this paper provides a multi-baseline PolInSAR forest height inversion scheme for exploring regional high-precision forest height distribution. The scheme is an applicable method for large-scale, high-precision forest height inversion tasks.

Deep learning for retrieving omni-directional ocean wave spectra from spaceborne synthetic aperture radar

Synthetic Aperture Radar (SAR) is a unique remote sensing instrument imaging ocean surface waves in two dimensions with high spatial resolution regardless of sunlight and weather conditions. However, due to the nonlinear imaging process, the ocean wave spectra cannot be retrieved directly from SAR data. The emergence of deep learning (DL) techniques provides a new paradigm for addressing this challenge. In this paper, a deep-learning-based model, called Wave-Spec-CNN, is proposed for retrieving omni-directional ocean wave spectra from SAR data. This model is constructed using approximately 21,000 collocations of Sentinel-1 Interferometric Wide swath mode images matched with global in-situ buoy data. The model adapts the convolution neural network (CNN) to accommodate the multi-valued nature of omni-directional ocean wave spectra, enhances performance by integrating a calibration branch and further incorporates physical characteristics into the training process. The results demonstrate consistency with buoy measurements for significant wave height (SWH) in the range of 0.5 m to 6 m, yielding a root-mean-square error (RMSE) of 0.51 m on the validation dataset, comparable to traditional physical-based methods. In terms of mean wave period (MWP) and peak frequency (PF), the achieved RMSEs are of 1.24 s and 0.03 Hz, respectively. The retrieved omni-directional ocean wave spectra also allow to separate swell and windsea components for respective comparisons with those derived by in-situ buoy data. The RMSEs of respective SWH comparisons are of 0.46 m and 0.42 m. This research represents an initial endeavor into utilizing DL for the long-standing challenge of SAR inversion for ocean wave spectra, as well as providing valuable insights for employing DL in multi-parameter inversion tasks in remote sensing.

Reconstruction of excitation waves from mechanical deformation using physics-informed neural networks

Non-invasive assessment of the electrical activation pattern can significantly contribute to the diagnosis and treatment of cardiac arrhythmias, due to faster and safer diagnosis, improved surgical planning and easier follow-up. One promising path is to measure the mechanical contraction via echocardiography and utilize this as an indirect way of measuring the original activation pattern. To solve this demanding inversion task, we make use of physics-informed neural networks, an upcoming methodology to solve forward and inverse physical problems governed by partial differential equations. In this study, synthetic data sets were created, consisting of 2D excitation waves coupled to an isotropic and linearly deforming elastic medium. We show that for both focal and spiral patterns, the underlying excitation waves can be reconstructed accurately. We test the robustness of the method against Gaussian noise, reduced spatial resolution and projected tri-planar data. In situations where the data quality is heavily reduced, we show how to improve the reconstruction by additional regularization on the wave speed. Our findings suggest that physics-informed neural networks hold the potential to solve sparse and noisy bio-mechanical inversion problems and may offer a pathway to non-invasive assessment of certain cardiac arrhythmias.

Predicting multiphase flow behavior of methane in shallow unconfined aquifers using conditional deep convolutional generative adversarial network

Numerical modeling is an essential tool for geoscience applications involving multiphase flow behavior. Performing numerical simulation is, however, computationally intensive due to multi-scale, non-linear, and multi-physics nature of mechanisms controlling multiphase flow dynamics. Additionally, the computational challenges associated with traditional numerical simulations limit their applicability for repetitive simulations required in inverse modeling, uncertainty analysis, and optimization tasks. To overcome these limitations, surrogate modeling has recently emerged as a viable solution. A surrogate model, which functions as a regression model, is trained using numerical simulation data. It approximates the input–output relationships within a dynamic fluid environment. We design surrogate models, based on conditional deep convolutional generative adversarial network (cDC-GAN), to map the cross-domain between input and output pairs in a multiphase multicomponent system, focusing on methane (CH4) migration in shallow unconfined aquifers. The cDC-GAN technique is selected due to its ability to generate high-fidelity surrogate models that effectively capture complex spatial distributions and heterogeneities. This technique excels in handling high-dimensional data, learning the intricate relationships between inputs and outputs, and producing realistic representations of subsurface phenomena. The trained cDC-GANs consider capillary entry pressure heterogeneity as input because this geologic attribute places first-order control on CH4 migration in subsurface. The outputs include CH4 saturation, water saturation, residual saturation of CH4, CH4 mole fraction, and CH4 molality. For each output, a separate cDC-GAN model is trained. Once trained, cDC-GANs can at any given time determine CH4 distribution in a shallow unconfined aquifer, both qualitatively and quantitatively. The cDC-GAN approach can be considered as a valuable complement to numerical simulations for predicting multiphase flow behavior in other geoscience-, energy-, and environmental applications such as contaminant transport, hydrocarbon production, and geological carbon dioxide and hydrogen storage.

Solving high-dimensional parametric engineering problems for inviscid flow around airfoils based on physics-informed neural networks

Engineering problems often involve solving partial differential equations (PDEs) over a range of similar problem setups, leading to high computational costs when using traditional numerical approaches to solve each setup individually. Recently developed physics-informed neural networks (PINNs) offer a novel approach to solving parametric problems, enabling the simultaneous solution of a series of similar problems. Our previous research combined PINNs with mesh transformation to learn PDE solutions in the computational space, effectively solving inviscid flow around airfoils. In this study, we expand the input dimensions of the model to include shape parameters and flow conditions, forming an input encompassing the complete state-space (i.e., all parameters determining the solution are included in the input). We spend about 18.8 h achieving the continuous solutions in a large state-space in one go, encompassing various subsonic inviscid airfoil flows encountered in engineering, thereby highlighting the model's significant advantages in addressing high-dimensional parametric problems. Once established, the model can efficiently complete airfoil flow simulation and shape inverse design tasks in about 1 s. Furthermore, we introduce a pretraining-finetuning method, enabling the fine-tuning of the model for the task of interest and quickly achieving accuracy comparable to the finite volume method.

Aligning the domains in cross domain model inversion attack

Model Inversion Attack reconstructs confidential training dataset from a target deep learning model. Most of the existing methods assume the adversary has an auxiliary dataset that has similar distribution with the private dataset. However, this assumption does not always hold in real-world scenarios. Since the private dataset is unknown, the domain divergence between the auxiliary dataset and the private dataset is inevitable. In this paper, we use Cross Domain Model Inversion Attack to represent the distribution divergence scenario in MIA. With the distribution divergence between the private images and auxiliary images, the distribution between the feature vectors of the private images and those of the auxiliary images is also different. Moreover, the outputted prediction vectors of the auxiliary images are also misclassified. The inversion attack is thus hard to be performed. We perform both the feature vector inversion task and prediction vector inversion task in this cross domain setting. For feature vector inversion, Domain Alignment MIA (DA-MIA) is proposed. While performing the reconstruction task, DA-MIA aligns the feature vectors of auxiliary images with the feature vectors of private images in an adversarial manner to mitigate the domain divergence between them. Thus, semantically meaningful images can be reconstructed. For prediction vector inversion, we further introduce an auxiliary classifier and propose Domain Alignment MIA with Auxiliary Classifier (DA-MIA-AC). The auxiliary classifier is pretrained by the auxiliary dataset and fine-tuned during the adversarial training stage. Thus, the misclassification problem caused by domain divergence can be solved, and the images can be reconstructed correctly. Various experiments are performed to show the advancement of our methods, the results show that DA-MIA can improve the SSIM score of the reconstructed images for up to 191%, DA-MIA-AC can increase the classification accuracy score of the reconstructed images from 9.18% to 81.32% in Cross Domain Model Inversion Attack.

Zadanie lokalizacji w sieciach kanalizacyjnych

Paper deals with the inverse / localisation task in sewer networks. An inverse problem is defined as the process of determining the causal factors from a set of observations. Applying this principle to the water management sector, it is often a matter of determining the location of the source of pollution based on monitored data on the concentration of pollution over time. From a mathematical point of view, to decrease the uncertainty of the inverse task solution, it is necessary to know the location of the source or the concentrations time course of the source (intensity function). In practice, we usually do not know any of these quantities, however, in the case of sewer networks we can accept some assumptions, which allow us to solve this inverse problem. Paper analyses specific conditions applied in the environment of sewer networks and describes proposed method for solving the source localisation task in the sewer network environment. The solution is based on numerical modelling of the pollution spreading in sewer system, accepting some process simplifications as well as assuming some source parameters. Typically, the solution of inverse task requires large and time consuming numerical simulations. This can be disadvantageous after recording the pollution event - a long calculation time reduces the efficiency and operability for the following pollution source reconnaissance. Therefore, our proposed method performs the necessary simulations in advance and the pollution source localisation after recording the pollution event is very fast, using a simple search and comparison in the simulation results database. The proposed method was tested on real sewer system and there were achieving promising results.

The heterogeneity of holistic processing profiles in developmental prosopagnosia: holistic processing is impaired but not absent

ABSTRACT Although it is generally assumed that face recognition relies on holistic processing, whether face recognition deficits observed in Developmental Prosopagnosics (DPs) can be explained by impaired holistic processing is currently under debate. The mixed findings from past studies could be the consequence of DP’s heterogeneous deficit nature and the use of different measures of holistic processing—the inversion, part-whole, and composite tasks—which showed a poor association among each other. The present study aimed to gain further insight into the role of holistic processing in DPs. Groups of DPs and neurotypicals completed three tests measuring holistic face processing and non-face objects (i.e., Navon task). At a group level, DPs showed (1) diminished, but not absent, inversion and part-whole effects, (2) comparable magnitudes of the composite face effect and (3) global precedence effect in the Navon task. However, single-case analyses showed that these holistic processing deficits in DPs are heterogeneous.

Learning data distribution of three-dimensional ocean sound speed fields via diffusion models.

The probability distribution of three-dimensional sound speed fields (3D SSFs) in an ocean region encapsulates vital information about their variations, serving as valuable data-driven priors for SSF inversion tasks. However, learning such a distribution is challenging due to the high dimensionality and complexity of 3D SSFs. To tackle this challenge, we propose employing the diffusion model, a cutting-edge deep generative model that has showcased remarkable performance in diverse domains, including image and audio processing. Nonetheless, applying this approach to 3D ocean SSFs encounters two primary hurdles. First, the lack of publicly available well-crafted 3D SSF datasets impedes training and evaluation. Second, 3D SSF data consist of multiple 2D layers with varying variances, which can lead to uneven denoising during the reverse process. To surmount these obstacles, we introduce a novel 3D SSF dataset called 3DSSF, specifically designed for training and evaluating deep generative models. In addition, we devise a high-capacity neural architecture for the diffusion model to effectively handle variations in 3D sound speeds. Furthermore, we employ state-of-the-art continuous-time-based optimization method and predictor-corrector scheme for high-performance training and sampling. Notably, this paper presents the first evaluation of the diffusion model's effectiveness in generating 3D SSF data. Numerical experiments validate the proposed method's strong ability to learn the underlying data distribution of 3D SSFs, and highlight its effectiveness in assisting SSF inversion tasks and subsequently characterizing the transmission loss of underwater acoustics.

TECHNOLOGY OF THE DEVELOPMENT OF MACHINE-GRAPHICS PROFILE OF A PARTS AND ASSEMBLY UNITS IN TRAINING SKETCHES

The purpose of studying the discipline “Engineering Graphics” is to acquire skills in working with engineering and graphic information: mapping spatial objects to orthogonal planes in the form of a drawing (direct task), reading a drawing (inverse task), ultimately solving spatial and logical problems. The necessity to start performing graphic works with a pencil, paper, using drawing and graphic tools is shown and justified, and only then to involve modern computer technologies for these purposes. Not the ability to work with a pencil; it is not understood that the use of computer graphics hardware and software at the initial stage leads to the loss of graphic competence of the graduate, and ultimately to a decrease in the competitive ability of specialists.

Spatial-Contextual Discrepancy Information Compensation for GAN Inversion

Most existing GAN inversion methods either achieve accurate reconstruction but lack editability or offer strong editability at the cost of fidelity. Hence, how to balance the distortion-editability trade-off is a significant challenge for GAN inversion. To address this challenge, we introduce a novel spatial-contextual discrepancy information compensation-based GAN-inversion method (SDIC), which consists of a discrepancy information prediction network (DIPN) and a discrepancy information compensation network (DICN). SDIC follows a ``compensate-and-edit'' paradigm and successfully bridges the gap in image details between the original image and the reconstructed/edited image. On the one hand, DIPN encodes the multi-level spatial-contextual information of the original and initial reconstructed images and then predicts a spatial-contextual guided discrepancy map with two hourglass modules. In this way, a reliable discrepancy map that models the contextual relationship and captures fine-grained image details is learned. On the other hand, DICN incorporates the predicted discrepancy information into both the latent code and the GAN generator with different transformations, generating high-quality reconstructed/edited images. This effectively compensates for the loss of image details during GAN inversion. Both quantitative and qualitative experiments demonstrate that our proposed method achieves the excellent distortion-editability trade-off at a fast inference speed for both image inversion and editing tasks. Our code is available at https://github.com/ZzqLKED/SDIC.

Multi-Modal Convolutional Parameterisation Network for Guided Image Inverse Problems.

There are several image inverse tasks, such as inpainting or super-resolution, which can be solved using deep internal learning, a paradigm that involves employing deep neural networks to find a solution by learning from the sample itself rather than a dataset. For example, Deep Image Prior is a technique based on fitting a convolutional neural network to output the known parts of the image (such as non-inpainted regions or a low-resolution version of the image). However, this approach is not well adjusted for samples composed of multiple modalities. In some domains, such as satellite image processing, accommodating multi-modal representations could be beneficial or even essential. In this work, Multi-Modal Convolutional Parameterisation Network (MCPN) is proposed, where a convolutional neural network approximates shared information between multiple modes by combining a core shared network with modality-specific head networks. The results demonstrate that these approaches can significantly outperform the single-mode adoption of a convolutional parameterisation network on guided image inverse problems of inpainting and super-resolution.

Methodology for Studying Combustion of Solid Rocket Propellants using Artificial Neural Networks

The combustion properties of energetic materials have been extensively studied in the scientific literature. With the rapid advancement of data science and artificial intelligence techniques, predicting the performance of solid rocket propellants (SRPs) has become a key focus for researchers globally. Understanding and forecasting the characteristics of SRPs are crucial for analyzing and modeling combustion mechanisms, leading to the development of cutting-edge energetic materials. This study presents a methodology utilizing artificial neural networks (ANN) to create multifactor computational models (MCM) for predicting the burning rate of solid propellants. These models, based on existing burning rate data, can solve direct and inverse tasks, as well as conduct virtual experiments. The objective functions of the models focus on burning rate (direct tasks) and pressure (inverse tasks). This research lays the foundation for developing generalized combustion models to forecast the effects of various catalysts on a range of SRPs. Furthermore, this work represents a new direction in combustion science, contributing to the creation of a High-Energetic Materials Genome that accelerates the development of advanced propellants.

Social perception of robots is shaped by beliefs about their minds

Roboticists often imbue robots with human-like physical features to increase the likelihood that they are afforded benefits known to be associated with anthropomorphism. Similarly, deepfakes often employ computer-generated human faces to attempt to create convincing simulacra of actual humans. In the present work, we investigate whether perceivers’ higher-order beliefs about faces (i.e., whether they represent actual people or android robots) modulate the extent to which perceivers deploy face-typical processing for social stimuli. Past work has shown that perceivers’ recognition performance is more impacted by the inversion of faces than objects, thus highlighting that faces are processed holistically (i.e., as Gestalt), whereas objects engage feature-based processing. Here, we use an inversion task to examine whether face-typical processing is attenuated when actual human faces are labeled as non-human (i.e., android robot). This allows us to employ a task shown to be differentially sensitive to social (i.e., faces) and non-social (i.e., objects) stimuli while also randomly assigning face stimuli to seem real or fake. The results show smaller inversion effects when face stimuli were believed to represent android robots compared to when they were believed to represent humans. This suggests that robots strongly resembling humans may still fail to be perceived as “social” due pre-existing beliefs about their mechanistic nature. Theoretical and practical implications of this research are discussed.

Deep learning applied to CO2 power plant emissions quantification using simulated satellite images

Abstract. The quantification of emissions of greenhouse gases and air pollutants through the inversion of plumes in satellite images remains a complex problem that current methods can only assess with significant uncertainties. The anticipated launch of the CO2M (Copernicus Anthropogenic Carbon Dioxide Monitoring) satellite constellation in 2026 is expected to provide high-resolution images of CO2 (carbon dioxide) column-averaged mole fractions (XCO2), opening up new possibilities. However, the inversion of future CO2 plumes from CO2M will encounter various obstacles. A challenge is the low CO2 plume signal-to-noise ratio due to the variability in the background and instrumental errors in satellite measurements. Moreover, uncertainties in the transport and dispersion processes further complicate the inversion task. To address these challenges, deep learning techniques, such as neural networks, offer promising solutions for retrieving emissions from plumes in XCO2 images. Deep learning models can be trained to identify emissions from plume dynamics simulated using a transport model. It then becomes possible to extract relevant information from new plumes and predict their emissions. In this paper, we develop a strategy employing convolutional neural networks (CNNs) to estimate the emission fluxes from a plume in a pseudo-XCO2 image. Our dataset used to train and test such methods includes pseudo-images based on simulations of hourly XCO2, NO2 (nitrogen dioxide), and wind fields near various power plants in eastern Germany, tracing plumes from anthropogenic and biogenic sources. CNN models are trained to predict emissions from three power plants that exhibit diverse characteristics. The power plants used to assess the deep learning model's performance are not used to train the model. We find that the CNN model outperforms state-of-the-art plume inversion approaches, achieving highly accurate results with an absolute error about half of that of the cross-sectional flux method and an absolute relative error of ∼ 20 % when only the XCO2 and wind fields are used as inputs. Furthermore, we show that our estimations are only slightly affected by the absence of NO2 fields or a detection mechanism as additional information. Finally, interpretability techniques applied to our models confirm that the CNN automatically learns to identify the XCO2 plume and to assess emissions from the plume concentrations. These promising results suggest a high potential of CNNs in estimating local CO2 emissions from satellite images.

Investigating the impact of hyperspectral reconstruction techniques on the quantitative inversion of rice physiological parameters: A case study using the MST++ model

Quantitative inversion is a significant topic in remote sensing science. The development of visible light-based hyperspectral reconstruction techniques has opened up novel prospects for low-cost, high-precision remote sensing inversion in agriculture. The aim of this study was to assesses the effectiveness of hyperspectral reconstruction technology in agricultural remote sensing applications. Hyperspectral images were reconstructed using the MST++ hyperspectral reconstruction model and compared with the original visible light images in terms of their correlation with physiological parameters, the accuracy of single-feature modeling, and the accuracy of combined feature modeling. The results showed that compared to the visible light image, the reconstructed data exhibited a stronger correlation with physiological parameters, and the accuracy was improved in both the single-feature and the combined feature inversion mode. However, compared to multispectral sensors, hyperspectral reconstruction provided limited improvement on the inversion model accuracy. It was concluded that for physiological parameters that are not easy to be directly observed, deep mining of features in visible light. data through hyperspectral reconstruction technology can improve the accuracy of the inversion model. Appropriate feature selection and simple models are more suitable for the remote sensing inversion task of traditional agronomic plot experiments. To strengthen the application of hyperspectral reconstruction technology in agricultural remote sensing, further development is necessary with broader wavelength ranges and more diverse agricultural scenes.

Approximating Intermediate Feature Maps of Self-Supervised Convolution Neural Network to Learn Hard Positive Representations in Chest Radiography.

Recent advances in contrastive learning have significantly improved the performance of deep learning models. In contrastive learning of medical images, dealing with positive representation is sometimes difficult because some strong augmentation techniques can disrupt contrastive learning owing to the subtle differences between other standardized CXRs compared to augmented positive pairs; therefore, additional efforts are required. In this study, we propose intermediate feature approximation (IFA) loss, which improves the performance of contrastive convolutional neural networks by focusing more on positive representations of CXRs without additional augmentations. The IFA loss encourages the feature maps of a query image and its positive pair to resemble each other by maximizing the cosine similarity between the intermediate feature outputs of the original data and the positive pairs. Therefore, we used the InfoNCE loss, which is commonly used loss to address negative representations, and the IFA loss, which addresses positive representations, together to improve the contrastive network. We evaluated the performance of the network using various downstream tasks, including classification, object detection, and a generative adversarial network (GAN) inversion task. The downstream task results demonstrated that IFA loss can improve the performance of effectively overcoming data imbalance and data scarcity; furthermore, it can serve as a perceptual loss encoder for GAN inversion. In addition, we have made our model publicly available to facilitate access and encourage further research and collaboration in the field.

Forest Canopy Height Estimation by Integrating Structural Equation Modeling and Multiple Weighted Regression

As an important component of forest parameters, forest canopy height is of great significance to the study of forest carbon stocks and carbon cycle status. There is an increasing interest in obtaining large-scale forest canopy height quickly and accurately. Therefore, many studies have aimed to address this issue by proposing machine learning models that accurately invert forest canopy height. However, most of the these approaches feature PolSAR observations from a data-driven viewpoint in the feature selection part of the machine learning model, without taking into account the intrinsic mechanisms of PolSAR polarization observation variables. In this work, we evaluated the correlations between eight polarization observation variables, namely, T11, T22, T33, total backscattered power (SPAN), radar vegetation index (RVI), the surface scattering component (Ps), dihedral angle scattering component (Pd), and body scattering component (Pv) of Freeman-Durden three-component decomposition, and the height of the forest canopy. On this basis, a weighted inversion method for determining forest canopy height under the view of structural equation modeling was proposed. In this study, the direct and indirect contributions of the above eight polarization observation variables to the forest canopy height inversion task were estimated based on structural equation modeling. Among them, the indirect contributions were generated by the interactions between the variables and ultimately had an impact on the forest canopy height inversion. In this study, the covariance matrix between polarization variables and forest canopy height was calculated based on structural equation modeling, the weights of the variables were calculated by combining with the Mahalanobis distance, and the weighted inversion of forest canopy height was carried out using PSO-SVR. In this study, some experiments were carried out using three Gaofen-3 satellite (GF-3) images and ICESat-2 forest canopy height data for some forest areas of Gaofeng Ridge, Baisha Lizu Autonomous County, Hainan Province, China. The results showed that T11, T33, and total backscattered power (SPAN) are highly correlated with forest canopy height. In addition, this study showed that determining the weights of different polarization observation variables contributes positively to the accurate estimation of forest canopy height. The forest canopy height-weighted inversion method proposed in this paper was shown to be superior to the multiple regression model, with a 26% improvement in r and a 0.88 m reduction in the root-mean-square error (RMSE).

Autonomous Output‐Oriented Aerosol Jet Printing Enabled by Hybrid Machine Learning

AbstractAdditive manufacturing (AM) is rapidly revolutionizing modern manufacturing with recent progress in advanced printing methods and improved properties of printed materials. However, traditional AM methods are limited by their input‐oriented nature, which demands tedious trial‐and‐error tuning of printing parameters to achieve desired output properties. Here, an output‐oriented artificial intelligence‐integrated AM (AIAM) method is reported that enables an user to specify desired output properties while the printer autonomously discovers the optimal input printing parameters by integrating hybrid machine learning models and in situ measurements. Based on a predictive mapping between the input printing parameters and the output properties of interests established with <20 experiments designed by active learning, inverse design tasks are performed to intelligently generate the printing parameter settings that lead to desired outcomes using reinforcement learning. This method is demonstrated by autonomous aerosol jet printing (AJP) of conductive polymer films and achieving user‐defined electrical resistances with an ultralow error of 3.7%. The AIAM method, with its output‐oriented nature, holds the potential to significantly improve the autonomy, predictability, efficiency, and accessibility of the AM processes, which will unlock new possibilities in the autonomous and intelligent printing of a broad range of functional materials and devices.