Nathaniel Blanchard

Huma Jamil, Yajing Liu, Christina M. Cole et al.

Previous work has shown that a neural network with the rectified linear unit (ReLU) activation function leads to a convex polyhedral decomposition of the input space. These decompositions can be represented by a dual graph with vertices corresponding to polyhedra and edges corresponding to polyhedra sharing a facet, which is a subgraph of a Hamming graph. This paper illustrates how one can utilize the dual graph to detect and analyze adversarial attacks in the context of digital images. When an image passes through a network containing ReLU nodes, the firing or non-firing at a node can be encoded as a bit ($1$ for ReLU activation, $0$ for ReLU non-activation). The sequence of all bit activations identifies the image with a bit vector, which identifies it with a polyhedron in the decomposition and, in turn, identifies it with a vertex in the dual graph. We identify ReLU bits that are discriminators between non-adversarial and adversarial images and examine how well collections of these discriminators can ensemble vote to build an adversarial image detector. Specifically, we examine the similarities and differences of ReLU bit vectors for adversarial images, and their non-adversarial counterparts, using a pre-trained ResNet-50 architecture. While this paper focuses on adversarial digital images, ResNet-50 architecture, and the ReLU activation function, our methods extend to other network architectures, activation functions, and types of datasets.

Soumyadip Roy, Chaitanya Roygaga, Nathaniel Blanchard et al.

Athletes routinely undergo fitness evaluations to evaluate their training progress. Typically, these evaluations require a trained professional who utilizes specialized equipment like force plates. For the assessment, athletes perform drop and squat jumps, and key variables are measured, e.g. velocity, flight time, and time to stabilization, to name a few. However, amateur athletes may not have access to professionals or equipment that can provide these assessments. Here, we investigate the feasibility of estimating key variables using video recordings. We focus on jump velocity as a starting point because it is highly correlated with other key variables and is important for determining posture and lower-limb capacity. We find that velocity can be estimated with a high degree of precision across a range of athletes, with an average R-value of 0.71 (SD = 0.06).

Ibrahim Khebour, Kenneth Lai, Mariah Bradford et al.

Within Dialogue Modeling research in AI and NLP, considerable attention has been spent on ``dialogue state tracking'' (DST), which is the ability to update the representations of the speaker's needs at each turn in the dialogue by taking into account the past dialogue moves and history. Less studied but just as important to dialogue modeling, however, is ``common ground tracking'' (CGT), which identifies the shared belief space held by all of the participants in a task-oriented dialogue: the task-relevant propositions all participants accept as true. In this paper we present a method for automatically identifying the current set of shared beliefs and ``questions under discussion'' (QUDs) of a group with a shared goal. We annotate a dataset of multimodal interactions in a shared physical space with speech transcriptions, prosodic features, gestures, actions, and facets of collaboration, and operationalize these features for use in a deep neural model to predict moves toward construction of common ground. Model outputs cascade into a set of formal closure rules derived from situated evidence and belief axioms and update operations. We empirically assess the contribution of each feature type toward successful construction of common ground relative to ground truth, establishing a benchmark in this novel, challenging task.

Abhijnan Nath, Videep Venkatesha, Mariah Bradford et al.

Question-asking in collaborative dialogue has long been established as key to knowledge construction, both in internal and collaborative problem solving. In this work, we examine probing questions in collaborative dialogues: questions that explicitly elicit responses from the speaker's interlocutors. Specifically, we focus on modeling the causal relations that lead directly from utterances earlier in the dialogue to the emergence of the probing question. We model these relations using a novel graph-based framework of deliberation chains, and reframe the problem of constructing such chains as a coreference-style clustering problem. Our framework jointly models probing and causal utterances and the links between them, and we evaluate on two challenging collaborative task datasets: the Weights Task and DeliData. Our results demonstrate the effectiveness of our theoretically-grounded approach compared to both baselines and stronger coreference approaches, and establish a standard of performance in this novel task.

Ethan Seefried, Ran Eldegaway, Sanjay Das et al.

Decades of engineering drawings and technical records remain locked in legacy archives with inconsistent or missing metadata, making retrieval difficult and often manual. We present Blueprint, a layout-aware multimodal retrieval system designed for large-scale engineering repositories. Blueprint detects canonical drawing regions, applies region-restricted VLM-based OCR, normalizes identifiers (e.g., DWG, part, facility), and fuses lexical and dense retrieval with a lightweight region-level reranker. Deployed on ~770k unlabeled files, it automatically produces structured metadata suitable for cross-facility search. We evaluate Blueprint on a 5k-file benchmark with 350 expert-curated queries using pooled, graded (0/1/2) relevance judgments. Blueprint delivers a 10.1% absolute gain in Success@3 and an 18.9% relative improvement in nDCG@3 over the strongest vision-language baseline}, consistently outperforming across vision, text, and multimodal intents. Oracle ablations reveal substantial headroom under perfect region detection and OCR. We release all queries, runs, annotations, and code to facilitate reproducible evaluation on legacy engineering archives.

Abhijnan Nath, Huma Jamil, Shafiuddin Rehan Ahmed et al.

Event coreference resolution (ECR) is the task of determining whether distinct mentions of events within a multi-document corpus are actually linked to the same underlying occurrence. Images of the events can help facilitate resolution when language is ambiguous. Here, we propose a multimodal cross-document event coreference resolution method that integrates visual and textual cues with a simple linear map between vision and language models. As existing ECR benchmark datasets rarely provide images for all event mentions, we augment the popular ECB+ dataset with event-centric images scraped from the internet and generated using image diffusion models. We establish three methods that incorporate images and text for coreference: 1) a standard fused model with finetuning, 2) a novel linear mapping method without finetuning and 3) an ensembling approach based on splitting mention pairs by semantic and discourse-level difficulty. We evaluate on 2 datasets: the augmented ECB+, and AIDA Phase 1. Our ensemble systems using cross-modal linear mapping establish an upper limit (91.9 CoNLL F1) on ECB+ ECR performance given the preprocessing assumptions used, and establish a novel baseline on AIDA Phase 1. Our results demonstrate the utility of multimodal information in ECR for certain challenging coreference problems, and highlight a need for more multimodal resources in the coreference resolution space.

Sarath Sreedharan, Kelsey Sikes, Nathaniel Blanchard et al.

The role that highly curated knowledge, provided by domain experts, could play in creating effective tutoring systems is often overlooked within the AI for education community. In this paper, we highlight this topic by discussing two ways such highly curated expert knowledge could help in creating novel educational systems. First, we will look at how one could use explainable AI (XAI) techniques to automatically create lessons. Most existing XAI methods are primarily aimed at debugging AI systems. However, we will discuss how one could use expert specified rules about solving specific problems along with novel XAI techniques to automatically generate lessons that could be provided to learners. Secondly, we will see how an expert specified curriculum for learning a target concept can help develop adaptive tutoring systems, that can not only provide a better learning experience, but could also allow us to use more efficient algorithms to create these systems. Finally, we will highlight the importance of such methods using a case study of creating a tutoring system for pollinator identification, where such knowledge could easily be elicited from experts.

Mariah Bradford, Nikhil Krishnaswamy, Nathaniel Blanchard

Interruption plays a crucial role in collaborative learning, shaping group interactions and influencing knowledge construction. AI-driven support can assist teachers in monitoring these interactions. However, most previous work on interruption detection and interpretation has been conducted in single-conversation environments with relatively clean audio. AI agents deployed in classrooms for collaborative learning within small groups will need to contend with multiple concurrent conversations -- in this context, overlapping speech will be ubiquitous, and interruptions will need to be identified in other ways. In this work, we analyze interruption detection in single-conversation and multi-group dialogue settings. We then create a state-of-the-art method for interruption identification that is robust to overlapping speech, and thus could be deployed in classrooms. Further, our work highlights meaningful linguistic and prosodic information about how interruptions manifest in collaborative group interactions. Our investigation also paves the way for future works to account for the influence of overlapping speech from multiple groups when tracking group dialog.

Videep Venkatesha, Mariah Bradford, Nathaniel Blanchard

Collaborative Problem-Solving (CPS) markers capture key aspects of effective teamwork, such as staying on task, avoiding interruptions, and generating constructive ideas. An AI system that reliably detects these markers could help teachers identify when a group is struggling or demonstrating productive collaboration. Such a system requires an automated pipeline composed of multiple components. In this work, we evaluate how CPS detection is impacted by automating two critical components: transcription and speech segmentation. On the public Weights Task Dataset (WTD), we find CPS detection performance with automated transcription and segmentation methods is comparable to human-segmented and manually transcribed data; however, we find the automated segmentation methods reduces the number of utterances by 26.5%, impacting the the granularity of the data. We discuss the implications for developing AI-driven tools that support collaborative learning in classrooms.

Videep Venkatesha, Mary Cati Poulos, Christopher Steadman et al.

The onset of spontaneous thoughts are reflective of dynamic interactions between cognition, emotion, and attention. Typically, these experiences are studied through subjective appraisals that focus on their triggers, phenomenology, and emotional salience. In this work, we use linguistic signatures to investigate Deja Vu, Involuntary Autobiographical Memories and Unexpected Thoughts. Specifically, we analyze the inherent characteristics of the linguistic patterns in participant generated descriptions of these thought types. We show how, by positioning language as a window into spontaneous cognition, existing theories on these attentional states can be updated and reaffirmed. Our findings align with prior research, reinforcing that Deja Vu is a metacognitive experience characterized by abstract and spatial language, Involuntary Autobiographical Memories are rich in personal and emotionally significant detail, and Unexpected Thoughts are marked by unpredictability and cognitive disruption. This work is demonstrative of languages potential to reveal deeper insights into how internal spontaneous cognitive states manifest through expression.

Changsoo Jung, Sheikh Mannan, Jack Fitzgerald et al.

Interactive and spatially aware technologies are transforming educational frameworks, particularly in K-12 settings where hands-on exploration fosters deeper conceptual understanding. However, during collaborative tasks, existing systems often lack the ability to accurately capture real-world interactions between students and physical objects. This issue could be addressed with automatic 6D pose estimation, i.e., estimation of an object's position and orientation in 3D space from RGB images or videos. For collaborative groups that interact with physical objects, 6D pose estimates allow AI systems to relate objects and entities. As part of this work, we introduce FiboSB, a novel and challenging 6D pose video dataset featuring groups of three participants solving an interactive task featuring small hand-held cubes and a weight scale. This setup poses unique challenges for 6D pose because groups are holistically recorded from a distance in order to capture all participants -- this, coupled with the small size of the cubes, makes 6D pose estimation inherently non-trivial. We evaluated four state-of-the-art 6D pose estimation methods on FiboSB, exposing the limitations of current algorithms on collaborative group work. An error analysis of these methods reveals that the 6D pose methods' object detection modules fail. We address this by fine-tuning YOLO11-x for FiboSB, achieving an overall mAP_50 of 0.898. The dataset, benchmark results, and analysis of YOLO11-x errors presented here lay the groundwork for leveraging the estimation of 6D poses in difficult collaborative contexts.

Corbyn Terpstra, Ibrahim Khebour, Mariah Bradford et al.

Collaborative problem solving (CPS) in teams is tightly coupled with the creation of shared meaning between participants in a situated, collaborative task. In this work, we assess the quality of different utterance segmentation techniques as an aid in annotating CPS. We (1) manually transcribe utterances in a dataset of triads collaboratively solving a problem involving dialogue and physical object manipulation, (2) annotate collaborative moves according to these gold-standard transcripts, and then (3) apply these annotations to utterances that have been automatically segmented using toolkits from Google and OpenAI's Whisper. We show that the oracle utterances have minimal correspondence to automatically segmented speech, and that automatically segmented speech using different segmentation methods is also inconsistent. We also show that annotating automatically segmented speech has distinct implications compared with annotating oracle utterances--since most annotation schemes are designed for oracle cases, when annotating automatically-segmented utterances, annotators must invoke other information to make arbitrary judgments which other annotators may not replicate. We conclude with a discussion of how future annotation specs can account for these needs.

Huma Jamil, Yajing Liu, Turgay Caglar et al.

Researchers typically investigate neural network representations by examining activation outputs for one or more layers of a network. Here, we investigate the potential for ReLU activation patterns (encoded as bit vectors) to aid in understanding and interpreting the behavior of neural networks. We utilize Representational Dissimilarity Matrices (RDMs) to investigate the coherence of data within the embedding spaces of a deep neural network. From each layer of a network, we extract and utilize bit vectors to construct similarity scores between images. From these similarity scores, we build a similarity matrix for a collection of images drawn from 2 classes. We then apply Fiedler partitioning to the associated Laplacian matrix to separate the classes. Our results indicate, through bit vector representations, that the network continues to refine class detectability with the last ReLU layer achieving better than 95\% separation accuracy. Additionally, we demonstrate that bit vectors aid in adversarial image detection, again achieving over 95\% accuracy in separating adversarial and non-adversarial images using a simple classifier.

David McNeely-White, Ben Sattelberg, Nathaniel Blanchard et al.

We present evidence that many common convolutional neural networks (CNNs) trained for face verification learn functions that are nearly equivalent under rotation. More specifically, we demonstrate that one face verification model's embeddings (i.e. last-layer activations) can be compared directly to another model's embeddings after only a rotation or linear transformation, with little performance penalty. This finding is demonstrated using IJB-C 1:1 verification across the combinations of ten modern off-the-shelf CNN-based face verification models which vary in training dataset, CNN architecture, method of angular loss calculation, or some combination of the 3. These networks achieve a mean true accept rate of 0.96 at a false accept rate of 0.01. When instead evaluating embeddings generated from two CNNs, where one CNN's embeddings are mapped with a linear transformation, the mean true accept rate drops to 0.95 using the same verification paradigm. Restricting these linear maps to only perform rotation produces a mean true accept rate of 0.91. These mappings' existence suggests that a common representation is learned by models despite variation in training or structure. We discuss the broad implications a result like this has, including an example regarding face template security.

Ameni Trabelsi, Ross J. Beveridge, Nathaniel Blanchard

Autonomous driving consists of a multitude of interacting modules, where each module must contend with errors from the others. Typically, the motion prediction module depends upon a robust tracking system to capture each agent's past movement. In this work, we systematically explore the importance of the tracking module for the motion prediction task and ultimately conclude that the overall motion prediction performance is highly sensitive to the tracking module's imperfections. We explicitly compare models that use tracking information to models that do not across multiple scenarios and conditions. We find that the tracking information plays an essential role and improves motion prediction performance in noise-free conditions. However, in the presence of tracking noise, it can potentially affect the overall performance if not studied thoroughly. We thus argue practitioners should be mindful of noise when developing and testing motion/tracking modules, or that they should consider tracking free alternatives.

David McNeely-White, Benjamin Sattelberg, Nathaniel Blanchard et al.

CNN feature spaces can be linearly mapped and consequently are often interchangeable. This equivalence holds across variations in architectures, training datasets, and network tasks. Specifically, we mapped between 10 image-classification CNNs and between 4 facial-recognition CNNs. When image embeddings generated by one CNN are transformed into embeddings corresponding to the feature space of a second CNN trained on the same task, their respective image classification or face verification performance is largely preserved. For CNNs trained to the same classes and sharing a common backend-logit (soft-max) architecture, a linear-mapping may always be calculated directly from the backend layer weights. However, the case of a closed-set analysis with perfect knowledge of classifiers is limiting. Therefore, empirical methods of estimating mappings are presented for both the closed-set image classification task and the open-set task of face recognition. The results presented expose the essentially interchangeable nature of CNNs embeddings for two important and common recognition tasks. The implications are far-reaching, suggesting an underlying commonality between representations learned by networks designed and trained for a common task. One practical implication is that face embeddings from some commonly used CNNs can be compared using these mappings.

Ameni Trabelsi, Mohamed Chaabane, Nathaniel Blanchard et al.

In this paper, we present a novel, end-to-end 6D object pose estimation method that operates on RGB inputs. Our approach is composed of 2 main components: the first component classifies the objects in the input image and proposes an initial 6D pose estimate through a multi-task, CNN-based encoder/multi-decoder module. The second component, a refinement module, includes a renderer and a multi-attentional pose refinement network, which iteratively refines the estimated poses by utilizing both appearance features and flow vectors. Our refiner takes advantage of the hybrid representation of the initial pose estimates to predict the relative errors with respect to the target poses. It is further augmented by a spatial multi-attention block that emphasizes objects' discriminative feature parts. Experiments on three benchmarks for 6D pose estimation show that our proposed pipeline outperforms state-of-the-art RGB-based methods with competitive runtime performance.

Matt Gorbett, Nathaniel Blanchard

Neural networks are vulnerable to a wide range of erroneous inputs such as adversarial, corrupted, out-of-distribution, and misclassified examples. In this work, we train a linear SVM classifier to detect these four types of erroneous data using hidden and softmax feature vectors of pre-trained neural networks. Our results indicate that these faulty data types generally exhibit linearly separable activation properties from correct examples, giving us the ability to reject bad inputs with no extra training or overhead. We experimentally validate our findings across a diverse range of datasets, domains, pre-trained models, and adversarial attacks.

Mohamed Chaabane, Ameni Trabelsi, Nathaniel Blanchard et al.

In this paper, we present an end-to-end future-prediction model that focuses on pedestrian safety. Specifically, our model uses previous video frames, recorded from the perspective of the vehicle, to predict if a pedestrian will cross in front of the vehicle. The long term goal of this work is to design a fully autonomous system that acts and reacts as a defensive human driver would --- predicting future events and reacting to mitigate risk. We focus on pedestrian-vehicle interactions because of the high risk of harm to the pedestrian if their actions are miss-predicted. Our end-to-end model consists of two stages: the first stage is an encoder/decoder network that learns to predict future video frames. The second stage is a deep spatio-temporal network that utilizes the predicted frames of the first stage to predict the pedestrian's future action. Our system achieves state-of-the-art accuracy on pedestrian behavior prediction and future frames prediction on the Joint Attention for Autonomous Driving (JAAD) dataset.

Nathaniel Blanchard, Daniel Moreira, Aparna Bharati et al.

In the last decade, video blogs (vlogs) have become an extremely popular method through which people express sentiment. The ubiquitousness of these videos has increased the importance of multimodal fusion models, which incorporate video and audio features with traditional text features for automatic sentiment detection. Multimodal fusion offers a unique opportunity to build models that learn from the full depth of expression available to human viewers. In the detection of sentiment in these videos, acoustic and video features provide clarity to otherwise ambiguous transcripts. In this paper, we present a multimodal fusion model that exclusively uses high-level video and audio features to analyze spoken sentences for sentiment. We discard traditional transcription features in order to minimize human intervention and to maximize the deployability of our model on at-scale real-world data. We select high-level features for our model that have been successful in nonaffect domains in order to test their generalizability in the sentiment detection domain. We train and test our model on the newly released CMU Multimodal Opinion Sentiment and Emotion Intensity (CMUMOSEI) dataset, obtaining an F1 score of 0.8049 on the validation set and an F1 score of 0.6325 on the held-out challenge test set.

Nathaniel Blanchard, Jeffery Kinnison, Brandon RichardWebster et al.

Neuroscience theory posits that the brain's visual system coarsely identifies broad object categories via neural activation patterns, with similar objects producing similar neural responses. Artificial neural networks also have internal activation behavior in response to stimuli. We hypothesize that networks exhibiting brain-like activation behavior will demonstrate brain-like characteristics, e.g., stronger generalization capabilities. In this paper we introduce a human-model similarity (HMS) metric, which quantifies the similarity of human fMRI and network activation behavior. To calculate HMS, representational dissimilarity matrices (RDMs) are created as abstractions of activation behavior, measured by the correlations of activations to stimulus pairs. HMS is then the correlation between the fMRI RDM and the neural network RDM across all stimulus pairs. We test the metric on unsupervised predictive coding networks, which specifically model visual perception, and assess the metric for statistical significance over a large range of hyperparameters. Our experiments show that networks with increased human-model similarity are correlated with better performance on two computer vision tasks: next frame prediction and object matching accuracy. Further, HMS identifies networks with high performance on both tasks. An unexpected secondary finding is that the metric can be employed during training as an early-stopping mechanism.