As autonomous systems grow more advanced, objective metrics to evaluate their ethical and legal compliance are critical for informing end users of their limitations and ensuring accountability of those who misuse them. Current ethical embodied AI frameworks remain mostly qualitative, focusing on system design (through safety guardrails or targeted red teaming), and the realized guardrails often directly disallow unsafe behavior without providing the user with an override or interpretable reason. Instead, there is a need for computable metrics through rigorous testing that allow a user to determine the applicability of the system to the task. To address this gap, we introduce the Reference Ethical Benchmark for Autonomy Readiness (REBAR), a quantitative test and evaluation framework for autonomous systems. REBAR maps operating metrics into a computable Autonomy Readiness Level (ARL) rubric that can quantify ethical performance. Key innovations of the framework include a neuro-symbolic Large Language Model (LLM) approach to calculate and explain the ethical difficulty of scenarios, LLM-driven at-scale generation of test instances, and a versatile, photorealistic simulation environment. By evaluating white-box autonomy solutions through this rigorous testing pipeline, REBAR delivers an objective and repeatable benchmark score, bridging the gap between abstract principles and verifiable, accountable autonomy.

The deployment of decision-making AI agents presents a critical challenge in maintaining alignment with human values or guidelines while operating in complex, dynamic environments. Agents trained solely to achieve their objectives may adopt harmful behavior, exposing a key trade-off between maximizing the reward function and maintaining alignment. For pre-trained agents, ensuring alignment is particularly challenging, as retraining can be a costly and slow process. This is further complicated by the diverse and potentially conflicting attributes representing the ethical values for alignment. To address these challenges, we propose a test-time alignment technique based on model-guided policy shaping. Our method allows precise control over individual behavioral attributes, generalizes across diverse reinforcement learning (RL) environments, and facilitates a principled trade-off between ethical alignment and reward maximization without requiring agent retraining. We evaluate our approach using the MACHIAVELLI benchmark, which comprises 134 text-based game environments and thousands of annotated scenarios involving ethical decisions. The RL agents are first trained to maximize the reward in their respective games. At test time, we apply policy shaping via scenario-action attribute classifiers to ensure decision alignment with ethical attributes. We compare our approach against prior training-time methods and general-purpose agents, as well as study several types of ethical violations and power-seeking behavior. Our results demonstrate that test-time policy shaping provides an effective and scalable solution for mitigating unethical behavior across diverse environments and alignment attributes.

We investigate how to characterize semantic significance (SS) in detecting image manipulations (IMD) for media forensics. We introduce the Characterization of Seman-tic Impact for IMD (CSI-IMD) dataset, which focuses on localizing and evaluating the semantic impact of image manipulations to counter advanced generative techniques. Our evaluation of 10 state-of-the-art IMD and localization methods on CSI-IMD reveals key insights. Unlike existing datasets, CSI-IMD provides detailed semantic annotations beyond traditional manipulation masks, aiding in the development of new defensive strategies. The dataset features manipulations from advanced generation methods, offering various levels of semantic significance. It is divided into two parts: a gold-standard set of 1,000 manu-ally annotated manipulations with high-quality control, and an extended set of 500,000 automated manipulations for large-scale training and analysis. We also propose a new SS-focused task to assess the impact of semantically targeted manipulations. Our experiments show that current IMD methods struggle with manipulations created using stable diffusion, with TruFor and Cat-Net performing the best among those tested. The CSI-IMD dataset will become available at https://github.com/csiimd/csiimd.

Large language models (LLMs) are currently aligned using techniques such as reinforcement learning from human feedback (RLHF). However, these methods use scalar rewards that can only reflect user preferences on average. Pluralistic alignment instead seeks to capture diverse user preferences across a set of attributes, moving beyond just helpfulness and harmlessness. Toward this end, we propose a steerable pluralistic model based on few-shot comparative regression that can adapt to individual user preferences. Our approach leverages in-context learning and reasoning, grounded in a set of fine-grained attributes, to compare response options and make aligned choices. To evaluate our algorithm, we also propose two new steerable pluralistic benchmarks by adapting the Moral Integrity Corpus (MIC) and the HelpSteer2 datasets, demonstrating the applicability of our approach to value-aligned decision-making and reward modeling, respectively. Our few-shot comparative regression approach is interpretable and compatible with different attributes and LLMs, while outperforming multiple baseline and state-of-the-art methods. Our work provides new insights and research directions in pluralistic alignment, enabling a more fair and representative use of LLMs and advancing the state-of-the-art in ethical AI.

We introduce KAMERA: a comprehensive system for multi-camera, multi-spectral synchronization and real-time detection of seals and polar bears. Utilized in aerial surveys for ice-associated seals in the Bering, Chukchi, and Beaufort seas around Alaska, KAMERA provides up to an 80% reduction in dataset processing time over previous methods. Our rigorous calibration and hardware synchronization enable using multiple spectra for object detection. All collected data are annotated with metadata so they can be easily referenced later. All imagery and animal detections from a survey are mapped onto a world plane for accurate surveyed area estimates and quick assessment of survey results. We hope KAMERA will inspire other mapping and detection efforts in the scientific community, with all software, models, and schematics fully open-sourced.

The rapid advancement of conversational agents, particularly chatbots powered by Large Language Models (LLMs), poses a significant risk of social engineering (SE) attacks on social media platforms. SE detection in multi-turn, chat-based interactions is considerably more complex than single-instance detection due to the dynamic nature of these conversations. A critical factor in mitigating this threat is understanding the SE attack mechanisms through which SE attacks operate, specifically how attackers exploit vulnerabilities and how victims' personality traits contribute to their susceptibility. In this work, we propose an LLM-agentic framework, SE-VSim, to simulate SE attack mechanisms by generating multi-turn conversations. We model victim agents with varying personality traits to assess how psychological profiles influence susceptibility to manipulation. Using a dataset of over 1000 simulated conversations, we examine attack scenarios in which adversaries, posing as recruiters, funding agencies, and journalists, attempt to extract sensitive information. Based on this analysis, we present a proof of concept, SE-OmniGuard, to offer personalized protection to users by leveraging prior knowledge of the victims personality, evaluating attack strategies, and monitoring information exchanges in conversations to identify potential SE attempts.

Artificial intelligence (AI) models are often viewed as black boxes, due to the inability to explain how they make their decisions or determine whether they will be robust to new data. As such, model explainability and model robustness are often stated as two key desiderata needed for establishing justified trust in AI, particularly in high-stakes scenarios. Saliency maps, a form of visual explainable AI, can be used to indicate which parts of the input an AI model paid attention to during its decision process. Natural robustness testing can be used to assess an AI model's performance envelope, by characterizing model degradation across a set of realistic, operationally-relevant perturbations. While aspects of model explainability and model robustness have previously been studied independently, we seek to better understand their relationship through quantification of how saliency maps change due to various perturbations. Quantitative changes in the structure and quality of saliency maps may provide a mechanism for understanding model failure modes and edge cases due to various perturbations. Focusing on the computer vision domain, we leverage the open source Explainable AI Toolkit (XAITK) and Natural Robustness Toolkit (NRTK) to perform experiments testing the relationship between saliency map quality and model natural robustness. We demonstrate results across several tasks, datasets, and models, covering various use cases. Finally, we relate this new framework for combined model explainability and robustness to current AI test and evaluation workflows.

Large language models (LLMs) are increasingly being used as decision aids. However, users have diverse values and preferences that can affect their decision-making, which requires novel methods for LLM alignment and personalization. Existing LLM comparison tools largely focus on benchmarking tasks, such as knowledge-based question answering. In contrast, our proposed ALIGN system focuses on dynamic personalization of LLM-based decision-makers through prompt-based alignment to a set of fine-grained attributes. Key features of our system include robust configuration management, structured output generation with reasoning, and several algorithm implementations with swappable LLM backbones, enabling different types of analyses. Our user interface enables a qualitative, side-by-side comparison of LLMs and their alignment to various attributes, with a modular backend for easy algorithm integration. Additionally, we perform a quantitative analysis comparing alignment approaches in two different domains: demographic alignment for public opinion surveys and value alignment for medical triage decision-making. The entire ALIGN framework is open source and will enable new research on reliable, responsible, and personalized LLM-based decision-makers.

Synthetically-generated imagery holds the promise of being a panacea for the challenges of real world datasets. Yet it continues to be frequently observed that deep learning model performance is not as good when trained with synthetic data versus real measured imagery. In this study we present analyses and illustration of the use of several statistical metrics, measures, and visualization tools based on the distance and similarity between real and synthetic data empirical distributions in the latent feature embedding space, which provide a quantitative understanding of the relevant image-domain distribution discrepancy issues hampering the generation of performant simulated datasets. We also demonstrate the practical applications of these tools and techniques in a novel study comparing latent space embedding vector distributions of real, pristine synthetic, and synthetic modified by physics-based degradation models. The results may assist deep learning practitioners and synthetic imagery modelers with evaluating latent space embedding distributional dissimilarity and improving model performance when using simulation tools to generate synthetic imagery training data.

Tracking and classifying fish in optical underwater imagery presents several challenges which are encountered less frequently in terrestrial domains. Video may contain large schools comprised of many individuals, dynamic natural backgrounds, highly variable target scales, volatile collection conditions, and non-fish moving confusers including debris, marine snow, and other organisms. Additionally, there is a lack of large public datasets for algorithm evaluation available in this domain. The contributions of this paper is three fold. First, we present the FishTrack23 dataset which provides a large quantity of expert-annotated fish groundtruth tracks, in imagery and video collected across a range of different backgrounds, locations, collection conditions, and organizations. Approximately 850k bounding boxes across 26k tracks are included in the release of the ensemble, with potential for future growth in later releases. Second, we evaluate improvements upon baseline object detectors, trackers and classifiers on the dataset. Lastly, we integrate these methods into web and desktop interfaces to expedite annotation generation on new datasets.

The proliferation of Large Language Models (LLMs) poses challenges in detecting and mitigating digital deception, as these models can emulate human conversational patterns and facilitate chat-based social engineering (CSE) attacks. This study investigates the dual capabilities of LLMs as both facilitators and defenders against CSE threats. We develop a novel dataset, SEConvo, simulating CSE scenarios in academic and recruitment contexts, and designed to examine how LLMs can be exploited in these situations. Our findings reveal that, while off-the-shelf LLMs generate high-quality CSE content, their detection capabilities are suboptimal, leading to increased operational costs for defense. In response, we propose ConvoSentinel, a modular defense pipeline that improves detection at both the message and the conversation levels, offering enhanced adaptability and cost-effectiveness. The retrieval-augmented module in ConvoSentinel identifies malicious intent by comparing messages to a database of similar conversations, enhancing CSE detection at all stages. Our study highlights the need for advanced strategies to leverage LLMs in cybersecurity.

Learning from multiple sensors is challenging due to spatio-temporal misalignment and differences in resolution and captured spectra. To that end, we introduce GeoWATCH, a flexible framework for training models on long sequences of satellite images sourced from multiple sensor platforms, which is designed to handle image classification, activity recognition, object detection, or object tracking tasks. Our system includes a novel partial weight loading mechanism based on sub-graph isomorphism which allows for continually training and modifying a network over many training cycles. This has allowed us to train a lineage of models over a long period of time, which we have observed has improved performance as we adjust configurations while maintaining a core backbone.

We identify a bias in a commonly used dataset for building damage detection, evaluate its effects on existing deep learning models, and devise mitigation strategies to overcome it. We find that the data contains significantly more groups of damaged buildings than single ones leading to skewed machine learning evaluations. Consequently, deep learning models heavily rely on surrounding context rather than individual building damage when classifying supporting our claim. Specifically, the dataset includes extraneous damage surrounding buildings such as debris, fallen trees, and other damaged buildings which results in deep neural networks overfitting to these features. We analyze the top-5 solutions of the xView2 challenge, which focuses on building damage classification using satellite imagery as provided by the xBD dataset. Our experiments reveal that these models struggle to accurately identify isolated damaged buildings, potentially causing oversights in critical disaster scenarios and delaying humanitarian aid. Finally, we devise a new augmentation strategy to reduce this bias in disaster datasets and show it improves real-world outcomes.

Both few-shot learning and domain adaptation sub-fields in Computer Vision have seen significant recent progress in terms of the availability of state-of-the-art algorithms and datasets. Frameworks have been developed for each sub-field; however, building a common system or framework that combines both is something that has not been explored. As part of our research, we present the first unified framework that combines domain adaptation for the few-shot learning setting across 3 different tasks - image classification, object detection and video classification. Our framework is highly modular with the capability to support few-shot learning with/without the inclusion of domain adaptation depending on the algorithm. Furthermore, the most important configurable feature of our framework is the on-the-fly setup for incremental n-shot tasks with the optional capability to configure the system to scale to a traditional many-shot task. With more focus on Self-Supervised Learning (SSL) for current few-shot learning approaches, our system also supports multiple SSL pre-training configurations. To test our framework's capabilities, we provide benchmarks on a wide range of algorithms and datasets across different task and problem settings. The code is open source has been made publicly available here: https://gitlab.kitware.com/darpa_learn/learn

Most action recognition datasets and algorithms assume a closed world, where all test samples are instances of the known classes. In open set problems, test samples may be drawn from either known or unknown classes. Existing open set action recognition methods are typically based on extending closed set methods by adding post hoc analysis of classification scores or feature distances and do not capture the relations among all the video clip elements. Our approach uses the reconstruction error to determine the novelty of the video since unknown classes are harder to put back together and thus have a higher reconstruction error than videos from known classes. We refer to our solution to the open set action recognition problem as "Humpty Dumpty", due to its reconstruction abilities. Humpty Dumpty is a novel graph-based autoencoder that accounts for contextual and semantic relations among the clip pieces for improved reconstruction. A larger reconstruction error leads to an increased likelihood that the action can not be reconstructed, i.e., can not put Humpty Dumpty back together again, indicating that the action has never been seen before and is novel/unknown. Extensive experiments are performed on two publicly available action recognition datasets including HMDB-51 and UCF-101, showing the state-of-the-art performance for open set action recognition.

In this paper, we present the Multi-view Extended Videos with Identities (MEVID) dataset for large-scale, video person re-identification (ReID) in the wild. To our knowledge, MEVID represents the most-varied video person ReID dataset, spanning an extensive indoor and outdoor environment across nine unique dates in a 73-day window, various camera viewpoints, and entity clothing changes. Specifically, we label the identities of 158 unique people wearing 598 outfits taken from 8, 092 tracklets, average length of about 590 frames, seen in 33 camera views from the very large-scale MEVA person activities dataset. While other datasets have more unique identities, MEVID emphasizes a richer set of information about each individual, such as: 4 outfits/identity vs. 2 outfits/identity in CCVID, 33 viewpoints across 17 locations vs. 6 in 5 simulated locations for MTA, and 10 million frames vs. 3 million for LS-VID. Being based on the MEVA video dataset, we also inherit data that is intentionally demographically balanced to the continental United States. To accelerate the annotation process, we developed a semi-automatic annotation framework and GUI that combines state-of-the-art real-time models for object detection, pose estimation, person ReID, and multi-object tracking. We evaluate several state-of-the-art methods on MEVID challenge problems and comprehensively quantify their robustness in terms of changes of outfit, scale, and background location. Our quantitative analysis on the realistic, unique aspects of MEVID shows that there are significant remaining challenges in video person ReID and indicates important directions for future research.

Existing methods for open set action recognition focus on novelty detection that assumes video clips show a single action, which is unrealistic in the real world. We propose a new method for open set action recognition and novelty detection via MUlti-Label Evidential learning (MULE), that goes beyond previous novel action detection methods by addressing the more general problems of single or multiple actors in the same scene, with simultaneous action(s) by any actor. Our Beta Evidential Neural Network estimates multi-action uncertainty with Beta densities based on actor-context-object relation representations. An evidence debiasing constraint is added to the objective function for optimization to reduce the static bias of video representations, which can incorrectly correlate predictions and static cues. We develop a primal-dual average scheme update-based learning algorithm to optimize the proposed problem and provide corresponding theoretical analysis. Besides, uncertainty and belief-based novelty estimation mechanisms are formulated to detect novel actions. Extensive experiments on two real-world video datasets show that our proposed approach achieves promising performance in single/multi-actor, single/multi-action settings. Our code and models are released at https://github.com/charliezhaoyinpeng/mule.

Novel object detection in remote sensing is challenging due to small objects, background clutter and open-set recognition. To discover and identify objects that are not within the set of known classes in training, we developed the extreme value theory-based novel object detection framework. Specifically, we first employed the state-of-the-art object detector to extract object detections. Then, the extreme value model (EVM) is trained based on the features of extracted detections. Thus our method can characterize the distribution of outliers in the known classes-based distributions to classify novel objects. If the novelty score is larger than the pre-set threshold, we assign this sample to novel classes; otherwise the sample is classified by the original object detector. To adapt to our task, we hold out a novel set of 18 of overall 60 classes in the xView dataset in satellite imagery. The experimental results on the xView dataset show the effectiveness of our proposed approach over traditional softmax thresholding.

Advances in artificial intelligence (AI) using techniques such as deep learning have fueled the recent progress in fields such as computer vision. However, these algorithms are still often viewed as "black boxes", which cannot easily explain how they arrived at their final output decisions. Saliency maps are one commonly used form of explainable AI (XAI), which indicate the input features an algorithm paid attention to during its decision process. Here, we introduce the open source xaitk-saliency package, an XAI framework and toolkit for saliency. We demonstrate its modular and flexible nature by highlighting two example use cases for saliency maps: (1) object detection model comparison and (2) doppelganger saliency for person re-identification. We also show how the xaitk-saliency package can be paired with visualization tools to support the interactive exploration of saliency maps. Our results suggest that saliency maps may play a critical role in the verification and validation of AI models, ensuring their trusted use and deployment. The code is publicly available at: https://github.com/xaitk/xaitk-saliency.