Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “multi-modal machine learning”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

54 records · Page 3
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Coincident learning for unsupervised anomaly detection of scientific instruments

Abstract Anomaly detection is an important task for complex scientific experiments and other complex systems (e.g. industrial facilities, manufacturing), where failures in a sub-system can lead to lost data, poor performance, or even damage to components. While scientific facilities generate a wealth of data, labeled anomalies may be rare (or even nonexistent), and expensive to acquire. Unsupervised approaches are therefore common and typically search for anomalies either by distance or density of examples in the input feature space (or some associated low-dimensional representation). This paper presents a novel approach called coincident learning for anomaly detection (CoAD), which is specifically designed for multi-modal tasks and identifies anomalies based on coincident behavior across two different slices of the feature space. We define an unsupervised metric, F ^ β , out of analogy to the supervised classification F β statistic. CoAD uses F ^ β to train an anomaly detection algorithm on unlabeled data , based on the expectation that anomalous behavior in one feature slice is coincident with anomalous behavior in the other. The method is illustrated using a synthetic outlier data set and a MNIST-based image data set, and is compared to prior state-of-the-art on two real-world tasks: a metal milling data set and our motivating task of identifying RF station anomalies in a particle accelerator.

43 PARTICLE ACCELERATORS↗
DOE OSTI2020

Performance Evaluation of Augmented Teleoperation of Contact Manipulation Tasks

At Argonne National Laboratory, an enhanced telerobotic concept has been suggested by providing augmented virtual fixtures for precise and dexterous manipulation. The proposed technology incorporated a human-robot interface that can assist personnel on the processes of Deactivation and Decommissioning in sites of high radiation. It has been implemented on a dual-arm robot system, known as Baxter, by integrating the Robotic Operating System (ROS) with 3D sensors and visual-haptic operator interface. However, what remained before its practical utilization is the capability for precisely placing the virtual fixtures. To this end, the objectives involved camera calibration, which required hardware fabrication and setup, as well as modification and implementation of ROS. The goal of this Research and Development is to develop an augmented-reality- based operator interface for human-inclusive, intelligent, and autonomous robotic system for the Deactivation and Decommissioning (D and D) of radioactive sites. The objectives are to develop: Collaborative Robot; Multi-modal Mixed-Reality Human-Robot Interface; Intelligent and Autonomous Behaviors. In order to achieve this, Baxter's Head-Kinect camera and Left-Hand camera need to be calibrated to overlap the real environment with the simulated environment. Camera calibration, which is crucial for precisely placing the virtual fixtures, was successful after running the simulations. The real and simulated environments overlapped, permitting the utilization of augmented teleoperation and the dexterous manipulation of Baxter for D and D applications in nuclear sites. Future steps: Performance evaluation of augmented teleoperation of contact manipulation tasks: Multiple-Camera 3D Reconstruction and Automation. Oculus Interface: Augmented reality and machine learning for enhanced telerobotic operation (Augmented Tele-autonomy): Machine-learning techniques for complex and flexible autonomous robotic behaviors. Interactive remote operation of robotic systems. Mesh cutting with simulation and real robot.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗
DOE OSTI2021

High-dimensional Data-driven Energy optimization for Multi-Modal Transit Agencies (HD-EMMA) (Final Technical Report)

Public bus transit services in the U.S. are responsible for at least 19.7 million metric tons of CO 2 emission annually. Electric vehicles (EVs) can have a much lower environmental impact than comparable internal combustion engine vehicles (ICEVs), especially in urban areas. Unfortunately, EVs are also much more expensive than ICEVs. As a result, many public transit agencies can afford only mixed fleets of transit vehicles, consisting of EVs, hybrids (HEVs), and ICEVs. Transit agencies that operate such mixed fleets of vehicles face a challenging optimization problem: these agencies need to decide which vehicles are assigned to serving which transit trips. Since the advantage of EVs over ICEVs varies depending on the route and time of day (e.g., the benefit of EVs is higher in slower traffic with frequent stops and lower on highways), the assignment can have a significant effect on energy use and, hence, environmental impact. Through this project, we have developed reference data about energy collections and constructed a set of machine learning models that can accurately predict the energy consumption for the whole fleet at the level of each trip. We have used these models to develop a scheduling and assignment strategy that can rotate the different vehicle types across the transit agencies’ routes. The optimization algorithm ensures that the vehicles are matched to trips considering weather patterns, expected congestion, and road gradients to minimize the overall energy usage. We list the key observations from our project for other practitioners below. Details are available in the report, and the list of source code and our publications are included in the appendix. 1. We have demonstrated the feasibility of collecting, merging and analyzing large volumes of high-resolution real-world telemetry data from a mixed vehicle fleet. To mitigate the inherent noise of the recorded GPS points, the team developed an algorithm that filters data and maps the points onto a street. The algorithm considers previous and subsequent location measurements and different characteristics of nearby streets to determine how likely the vehicle travels on them. Then, the team segmented the time series into disjoint contiguous samples based on adjacent road segments and repeated the outlier detection and removal. For each data point, the team added features corresponding to elevation changes within the samples, weather features, such as temperature, and traffic data, such as speed ratio between actual speed and free-flow speed. 2. We have developed two forms of machine learning models that be used to understand and analyze the energy operations of a mixed vehicle transit fleet. The micro prediction model provides estimates of instantaneous energy prediction for all types of buses (diesel, hybrid, and electric). Such a model is important in evaluating the energy impacts of real-time bus operation strategies, but it is challenging due to diversified driving cycles of transit buses. The model can help the drivers understand the impact of their driving behaviors and short-term congestions. The macro prediction models estimate average energy consumption across the whole trip considering the features: distance traveled, various road-type features, elevation change, day of the week, time of day, various weather features (temperature, humidity, etc.), and traffic features (speed ratio and jam factor). 3. We have demonstrated that it is possible to transfer the machine learning models we have developed in this project to other teams and cities by using inductive transfer learning. We also showed that the performance of the macro energy prediction models can be improved using a multi-task learning approach where the learning parameters are shared between the models being developed for different vehicle types. The advantage of this approach is improved learning performance as the models can exploit common spatio-temporal and environmental characteristics. 4. Finally, we have developed trip and vehicle assignment and scheduling algorithms that use the energy prediction models and develop a trip to vehicle type (diesel, electric, hybrid) assignment for the whole operation to reduce overall emissions and cost. We have shown through simulations that the proposed algorithms can save $\$$ 48,910 in energy costs and 175 metric tons of CO 2 emission annually for CARTA.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
DOE OSTI2024

Machine-Learning-Based Mapping and Modeling of Solar Energy with Ultra-High Spatiotemporal Granularity

Despite the rapid growth of solar energy, we still lack a dynamic, high-fidelity database that tracks the spatiotemporal variations of solar PVs and their associated infrastructures across different places at a spatially resolved scale. The absence of such data presents a barrier to various applications such as solar PV growth projection, solar energy integration, solar incentive design, and climate risk assessment. In this project, we aim to bridge this gap by developing AI-based algorithms to extract granular information about solar PV installations and their associated infrastructures (i.e., distribution grids) from widely available unstructured data like remote sensing images and street views. As a result, we have built the Solar Energy Atlas, a fine-grained, large-scale geospatial overlay of distributed solar PVs and distribution grids. On top of it, we have advanced the understanding of solar adoption and distribution grid vulnerability to climate-induced extremes. Our major contributions can be summarized as follow: (1) By developing new AI algorithms, we have built the most comprehensive solar PV spatiotemporal database covering the entire US. This is the first time we obtained the exact GPS locations, size, subtype, and installation year information for rooftop solar PVs across the US. This database can be used for solar PV growth projection, solar energy integration, solar energy policy analysis and design, and spatially-resolved climate risk assessment. (2) Leveraging this database, we have uncovered the socioeconomic driving factors that are correlated with earlier onset of solar adoption and higher saturated adoption levels. We have identified the heterogeneity in the effects of different types of financial incentives on solar adoption and provided implications for tailoring incentive design based on local income levels to promote equitable solar adoption. (3) We have developed a distribution grid GIS mapping algorithm which can obtain granular geospatial and topology information about distribution grids using multi-modal open data, reducing the dependency on hard-to-obtain smart meter data of conventional approaches. It shows effectiveness in both the U.S. and Sub-Saharan Africa. Using this algorithm, we have uncovered the non-uniform vulnerability of distribution grids to wildfires in California in the aspects of undergrounding protection and Distributed Energy Resources (DER) preparedness. This has provided important implications for improving the affordability and equity of grid adaptation approaches. (3) We have made our produced database publicly available and provided user-friendly interface to enable various stakeholders and the general public to interact with the data. We have also integrated the produced data into the Data Commons platform to enable the public to access the data and correlate it with other location-specific characteristics simply using natural language as queries. The impact of our project is three-fold: (1) New algorithms for mapping solar PVs and distribution grids across space and time, which are open source to facilitate researchers and industry; (2) New databases of solar PVs and distribution grids that have been made publicly available for engineering, social, and policy applications; (3) New understandings and actionable insights on the potential approaches to promoting solar adoption and reducing energy infrastructure vulnerabilities. In this report, we start by discussing the project background and motivation (section 5), followed by the overview of project objectives (section 6). Results and discussion for each task are presented in section 7. Significant accomplishments are summarized in section 8. This report will be concluded by discussing the paths forwards (section 9), products (section 10), and team roles (section 11).

14 SOLAR ENERGY↗
DOE OSTI2023

Machine Learning Enabled Sensor Fusion for In-Situ Defect Detection in Laser Powder Bed Fusion

Laser Powder Bed Fusion (L-PBF) Additive Manufacturing (AM) is among the metal 3D printing technologies most broadly adopted by the manufacturing industry. The current industry qualification paradigm for critical-application L-PBF parts relies heavily on expensive non-destructive inspection techniques such as X-Ray Computed Tomography (XCT), which significantly limits the use-cases of L-PBF. In situ monitoring of the process promises a less expensive alternative to ex situ testing, but existing sensor technologies and data analysis techniques struggle to detect sub-surface flaws (e.g., porosity and cracking) on production-scale L-PBF printers. RTX Technologies Research Center (RTRC) has licensed ORNL’s Peregrine software package – a printer- and camera-agnostic data analytics tool designed specifically for detecting process anomalies using in situ data collected during powder bed printing. The goal of this project was to feed temporally rich, multi-modal sensor data, including visible light, integrated near infrared (NIR), and spatially mapped co-axial melt pool thermal emission data into Peregrine to enable detection of subsurface flaws. XCT data was used as ground truth training data to allow Peregrine’s deep learning algorithms to recognize anomalies in these complex data streams in both test artifacts and industrially relevant geometries. Completion of this program has seen the successful implementation of multi-modal, multi-layer sensor data footprints for training of machine learning models in Peregrine. Flaws detected in XCT data have been successfully detected directly from this in situ data footprint, and initial analyses of the in situ probability-of-detection has been conducted, showing performance levels commensurate with traditional non-destructive evaluation (NDE) methods. The in situ monitoring methodology was then applied to an industrially relevant component that was using post-build NDE, highlighting the utility of the proposed method for hard-to-inspect AM components. As a direct result of this program, two journal manuscripts [1], [2] have been published in Additive Manufacturing, with additional manuscripts planned following program completion.

36 MATERIALS SCIENCE↗
DOE OSTI2024

Rapid detection of rare events from in situ X-ray diffraction data using machine learning

High-energy X-ray diffraction methods can non-destructively map the 3D microstructure and associated attributes of metallic polycrystalline engineering materials in their bulk form. These methods are often combined with external stimuli such as thermo-mechanical loading to take snapshots of the evolving microstructure and attributes over time. However, the extreme data volumes and the high costs of traditional data acquisition and reduction approaches pose a barrier to quickly extracting actionable insights and improving the temporal resolution of these snapshots. This article presents a fully automated technique capable of rapidly detecting the onset of plasticity in high-energy X-ray microscopy data. The technique is computationally faster by at least 50 times than the traditional approaches and works for data sets that are up to nine times sparser than a full data set. This new technique leverages self-supervised image representation learning and clustering to transform massive data sets into compact, semantic-rich representations of visually salient characteristics ( e.g. peak shapes). These characteristics can rapidly indicate anomalous events, such as changes in diffraction peak shapes. It is anticipated that this technique will provide just-in-time actionable information to drive smarter experiments that effectively deploy multi-modal X-ray diffraction methods spanning many decades of length scales.

Zheng, Weijian↗
DOE OSTI2025

Unveiling the Hidden Evolution of Crystal Defects and Disorder in Energy Materials

Control of point defects and disorder in functional thin films and 2D materials is critical to realizing their full potential in applications ranging from energy storage to advanced electronics. However, these phenomena are often poorly understood, difficult to characterize, and challenging to direct with precision. This presentation explores emerging multi-modal computer vision to decipher and predict order in materials across multiple length scales in the electron microscope, from the atomic to the nanoscale. By fusing data from diverse sources, these powerful models provide unprecedented insights into materials' lifecycles, enabling the control of defects and their associated properties at a fundamental level. This capability promises to transform materials design and accelerate the development of next-generation technologies.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

Multi-modal characterization and simulation of human epileptic circuitry

Temporal lobe epilepsy is the fourth most common neurological disorder, with about 40% of patients not responding to pharmacological treatment. Increased cellular loss is linked to disease severity and pathological phenotypes such as heightened seizure propensity. While the hippocampus is the target of therapeutic interventions, the impact of the disease at the cellular level remains unclear. Here, we show that hippocampal granule cells change with disease progression as measured in living, resected hippocampal tissue excised from patients with epilepsy. We show that granule cells increase excitability and shorten response latency while also enlarging in cellular volume and spine density. Single-nucleus RNA sequencing combined with simulations ascribes the changes to three conductances: BK, Cav2.2, and Kir2.1. In a network model, we show that these changes related to disease progression bring the circuit into a more excitable state, while reversing them produces a less excitable, “early-disease-like” state.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2026

Multi-modal dynamic radiography using short-pulse laser-generated probe beams

Radiography is an important tool for the interrogation of dynamic experiments in the fields of dynamic properties of materials, and in condensed matter, high explosive, and high-energy-density physics. Multi-modal radiography advances the hypothesis that combining the information delivered by multiple radiographic modalities can lead to more constrained (improved) “reconstruction” of the scene than can be obtained from a single probe. We identify four modalities: multi-probe, time sequence, multi-view, and multi-messenger. Multi-probe radiography is a promising candidate for a next-generation dynamic radiographic facility. High-energy X-rays are the most frequently used probe for dynamic radiography, although recent developments show the utility of proton (pRad), electron (eRad), and neutron probe beams. Because each probing species interacts with material in the radiographic scene through quantitatively different mechanisms, each returns independent information about the scene, which can add extra constraints to the reconstruction process. How to conduct detailed, quantitative “co-analysis” of multiple data streams remains an area of active research. Multi-beam, short-pulse, laser-generated probes offer sufficient dose, an appropriate spectrum, and appropriate spatio-temporal resolution to produce high-quality dynamic radiographs. This paper reports on technology development to advance the state of the art of multi-modal/multi-probe radiography and the pursuit of both deterministic and inferential (AI/ML assisted) co-analysis methodologies to produce more constrained reconstructions from multi-modal data.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2025

Object detection with deep learning for rare event search in the GADGET II TPC

In the pursuit of identifying rare two-particle events within the GADGET II Time Projection Chamber (TPC), this paper presents a comprehensive approach for leveraging Convolutional Neural Networks (CNNs) and various data processing methods. To address the inherent complexities of 3D TPC track reconstructions, the data is expressed in 2D projections and 1D quantities. This approach capitalizes on the diverse data modalities of the TPC, allowing for the efficient representation of the distinct features of the 3D events, with no loss in topology uniqueness. Additionally, it leverages the computational efficiency of 2D CNNs and benefits from the extensive availability of pre-trained models. Given the scarcity of real training data for the rare events of interest, simulated events are used to train the models to detect real events. To account for potential distribution shifts when predominantly depending on simulations, significant perturbations are embedded within the simulations. This produces a broad parameter space that works to account for potential physics parameter and detector response variations and uncertainties. These parameter-varied simulations are used to train sensitive 2D CNN object detectors. When combined with 1D histogram peak detection algorithms, this multi-modal detection framework is highly adept at identifying rare, two-particle events in data taken during experiment 21072 at the Facility for Rare Isotope Beams (FRIB), demonstrating a 100% recall for events of interest. Here, we present the methods and outcomes of our investigation and discuss the potential future applications of these techniques.

Convolutional neural network↗
DOE OSTI2026

Ripening of Rh Nanoparticle Catalysts in Reverse Water–Gas Shift via a Data-Driven Model Combining Physics, Theory, and Experiment

Degradation via sintering is an ongoing challenge that impedes the broad commercial success of supported metallic nanoparticle catalysts. To mitigate degradation via informed catalyst design and process operations, here we aim to disambiguate the underlying mechanisms of sintering by combining theory and experiment in a quantitative framework. While mechanistic sintering models exist, they only model a single sintering pathway, even though multiple sintering mechanisms can occur simultaneously or dominate at different stages of the process. Data-driven machine learning models have emerged as a means to represent complex processes through data regression. However, machine learning models have very large data needs and lack mechanistic insights due to their black-box encoding. To develop an interpretive model of catalyst degradation via sintering, we constructed a hybrid model combining mechanistic “physics-based” models and data-driven methods to obtain both reliable predictions and mechanistic insights regarding experimentally observed sintering phenomena. Focusing on nanoparticle sintering in the Rh–TiO 2 catalyst for the reverse water–gas shift (RWGS) reaction, the hybrid model couples a mechanistic term for Ostwald ripening with energy values calculated via density functional theory (DFT) with a parametric, data-driven discrepancy function term for unmodeled mechanisms. The hybrid model is trained using Bayesian inference with data collected from small-angle X-ray scattering (SAXS) in situ experiments wherein average nanoparticle diameter versus time was measured at three relevant operating temperatures. The calibrated hybrid model results show that an Ostwald ripening-only model parameterized with fixed DFT energies does not fully capture the time and temperature dependence of the SAXS-observed sintering kinetics, and that an additional functional contribution, or DFT energy calibration, is required to reconcile simulation and experiment. Analysis of the hybrid-model error confirms that the hybrid model outperforms both the purely mechanistic and purely data-driven alternatives in terms of expected predictive accuracy for time-evolving average particle sizes. Furthermore, the results support the hypothesis that the Ostwald ripening mechanism is less important for explaining the sintering phenomena as operating temperature increases under an assumed fixed DFT parameterization. This could be explained in one of two ways: either latent, unmodeled sintering mechanisms dominate at higher temperatures, or the DFT uncertainty increases with temperature. The proposed modeling approach directly links theory to experiments and simulations via a statistical hybrid modeling framework and can be extended to other catalytic systems to improve predictive models and mechanistic understanding.

Bayesian hybrid modeling↗
DOE OSTI2022

Preparing for the next pandemic via transfer learning from existing diseases with hierarchical multi-modal BERT: a study on COVID-19 outcome prediction

Abstract Developing prediction models for emerging infectious diseases from relatively small numbers of cases is a critical need for improving pandemic preparedness. Using COVID-19 as an exemplar, we propose a transfer learning methodology for developing predictive models from multi-modal electronic healthcare records by leveraging information from more prevalent diseases with shared clinical characteristics. Our novel hierarchical, multi-modal model ( $${\textsc {TransMED}}$$ T R A N S MED ) integrates baseline risk factors from the natural language processing of clinical notes at admission, time-series measurements of biomarkers obtained from laboratory tests, and discrete diagnostic, procedure and drug codes. We demonstrate the alignment of $${\textsc {TransMED}}$$ T R A N S MED ’s predictions with well-established clinical knowledge about COVID-19 through univariate and multivariate risk factor driven sub-cohort analysis. $${\textsc {TransMED}}$$ T R A N S MED ’s superior performance over state-of-the-art methods shows that leveraging patient data across modalities and transferring prior knowledge from similar disorders is critical for accurate prediction of patient outcomes, and this approach may serve as an important tool in the early response to future pandemics.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2022

Suppressing simulation bias in multi-modal data using transfer learning

Abstract Many problems in science and engineering require making predictions based on few observations. To build a robust predictive model, these sparse data may need to be augmented with simulated data, especially when the design space is multi-dimensional. Simulations, however, often suffer from an inherent bias. Estimation of this bias may be poorly constrained not only because of data sparsity, but also because traditional predictive models fit only one type of observed outputs, such as scalars or images, instead of all available output data modalities, which might have been acquired and simulated at great cost. To break this limitation and open up the path for multi-modal calibration, we propose to combine a novel, transfer learning technique for suppressing the bias with recent developments in deep learning, which allow building predictive models with multi-modal outputs. First, we train an initial neural network model on simulated data to learn important correlations between different output modalities and between simulation inputs and outputs. Then, the model is partially retrained, or transfer learned, to fit the experiments; a method that has never been implemented in this type of architecture. Using fewer than 10 inertial confinement fusion experiments for training, transfer learning systematically improves the simulation predictions while a simple output calibration, which we design as a baseline, makes the predictions worse. We also offer extensive cross-validation with real and carefully designed synthetic data. The method described in this paper can be applied to a wide range of problems that require transferring knowledge from simulations to the domain of experiments.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2022

XRF-ROI Finder: Machine Learning to Guide Region-of-Interest Scanning for X-ray Fluorescence Microscopy

The ROI-finder software is being developed for use by several Microscopy Group beamlines at Argonne National Laboratory, including 2-ID microprobes and 9-ID-B Bionanoprobe which use multi-scale scanning fluorescence microscopy to acquire elemental maps (multi-modal image data). Microscopy experiments require scan of samples at a coarse resolution followed by ROI identification using feature detection based on domain expertise. Finer resolution scans are then conducted based on identified ROI. The decision-making process based on domain expertise will be difficult to perform for faster data rates and much larger sampling volumes anticipated after APS-U necessitating the need for the ROI-finder software. The ROI- finder detects regions of interest through a continuous learning process, starting with a unsupervised representation learning and improving its recommendations through supervised learning and an interactive tool for user annotation. The scope of ongoing development efforts includes the integration of image registration module to correlate optical and X-ray images, extraction of feature morphology as well as elemental signatures in the image space and incorporation of beamtime streaming data by the scanning probe via EPICS.

CHOWDHURY, M. ARSHAD ZAHANGIR↗
DOE OSTI2023

Iterative multi-task learning and inference from seismic images

Seismic interpretation aims to extract quantitative and interpretable attributes from a seismic image produced using some migration method to inform characteristics of a subsurface reservoir or target of interest. Current paradigms for computing seismic attributes mostly rely on single-task algorithms. We develop an iterative, multi-task machine learning method to learn and infer multiple attributes from a seismic image. This method is composed of two stages: a multi-task inference stage and a multi-modal, multi-task refinement stage. The basic mechanism of this method is that we train a multi-task inference neural network (NN) to estimate a set of attributes, including a relative geological time (RGT), a denoised higher-resolution (DHR) seismic image, and multiple fault attributes (including probability, dip, and strike), from a low-resolution, noisy seismic image; then we input the inferred attributes to a multi-task refinement NN to enhance the raw inference results iteratively. The two multi-task NNs are trained separately based on synthetic seismic images and associated attributes generated by a geological modeling algorithm. The software we intend to release is a PyTorch implementation of this multi-task learning method for both 2D and 3D cases along with scripts to run the training/validation. The algorithm and software can be a useful tool for automatic seismic interpretation.

Gao, Kai↗
DOE OSTI2024

Hiding-in-Plain-Sight (HiPS) Attack on CLIP for Targetted Object Removal from Images

Machine learning models are known to be vulnerable to adversarial attacks, but prior works have mostly focused on single-modalities. With the rise of large multi-modal models (LMMs) like CLIP, which combine vision and language capabilities, new vulnerabilities have emerged. However, these multimodal targeted attacks aim to completely change the model's output to what the adversary wants. In many realistic scenarios, an adversary might seek to make only subtle modifications to the output, so that the changes go unnoticed by downstream models or even by humans. We introduce Hiding-in-Plain-Sight (HiPS) attacks, a novel class of adversarial attacks that subtly modifies model predictions by selectively concealing target object(s), as if the target object was absent from the scene. We propose two HiPS attack variants, HiPS-cls and HiPS-cap, and demonstrate their effectiveness in transferring to downstream image captioning models, such as CLIP-Cap, for targeted object removal from image captions.

Daw, Arka [ORNL] (ORCID:0009000633191271)↗