arXiv subjects

Hari Iyer

Publications and source records attributed to Hari Iyer.

7 recordsLinked to original sources

arXiv2025

Generative AI-Driven High-Fidelity Human Motion Simulation

Human motion simulation (HMS) supports cost-effective evaluation of worker behavior, safety, and productivity in industrial tasks. However, existing methods often suffer from low motion fidelity. This study introduces Generative-AI-Enabled HMS (G-AI-HMS), which integrates text-to-text and text-to-motion models to enhance simulation quality for physical tasks. G-AI-HMS tackles two key challenges: (1) translating task descriptions into motion-aware language using Large Language Models aligned with MotionGPT's training vocabulary, and (2) validating AI-enhanced motions against real human movements using computer vision. Posture estimation algorithms are applied to real-time videos to extract joint landmarks, and motion similarity metrics are used to compare them with AI-enhanced sequences. In a case study involving eight tasks, the AI-enhanced motions showed lower error than human created descriptions in most scenarios, performing better in six tasks based on spatial accuracy, four tasks based on alignment after pose normalization, and seven tasks based on overall temporal similarity. Statistical analysis showed that AI-enhanced prompts significantly (p $<$ 0.0001) reduced joint error and temporal misalignment while retaining comparable posture accuracy.

cs.AI↗

arXiv2025

Bayesian Forensic DNA Mixture Deconvolution Using a Novel String Similarity Measure

Mixture interpretation is a central challenge in forensic science, where evidence often contains contributions from multiple sources. In the context of DNA analysis, biological samples recovered from crime scenes may include genetic material from several individuals, necessitating robust statistical tools to assess whether a specific person of interest (POI) is among the contributors. Methods based on capillary electrophoresis (CE) are currently in use worldwide, but offer limited resolution in complex mixtures. Advancements in massively parallel sequencing (MPS) technologies provide a richer, more detailed representation of DNA mixtures, but require new analytical strategies to fully leverage this information. In this work, we present a Bayesian framework for evaluating whether a POIs DNA is present in an MPS-based forensic sample. The model accommodates known contributors, such as the victim, and uses a novel string edit distance to quantify similarity between observed alleles and sequencing artifacts. The resulting Bayes factors enable effective discrimination between samples that do and do not contain the POIs DNA, demonstrating strong performance in both hypothesis testing and classification settings.

stat.ME↗

arXiv2024

Computer-Vision-Enabled Worker Video Analysis for Motion Amount Quantification

The performance of physical workers is significantly influenced by the extent of their motions. However, monitoring and assessing these motions remains a challenge. Recent advancements have enabled in-situ video analysis for real-time observation of worker behaviors. This paper introduces a novel framework for tracking and quantifying upper and lower limb motions, issuing alerts when critical thresholds are reached. Using joint position data from posture estimation, the framework employs Hotelling's $T^2$ statistic to quantify and monitor motion amounts. A significant positive correlation was noted between motion warnings and the overall NASA Task Load Index (TLX) workload rating (\textit{r} = 0.218, \textit{p} = 0.0024). A supervised Random Forest model trained on the collected motion data was benchmarked against multiple datasets including UCF Sports Action and UCF50, and was found to effectively generalize across environments, identifying ergonomic risk patterns with accuracies up to 94\%.

cs.CV↗

arXiv2024

The Influence of Validation Data on Logical and Scientific Interpretations of Forensic Expert Opinions

Forensic experts use specialized training and knowledge to enable other members of the judicial system to make better informed and more just decisions. Factfinders, in particular, are tasked with judging how much weight to give to experts' reports and opinions. Many references describe assessing evidential weight from the perspective of a forensic expert. Some recognize that stakeholders are each responsible for evaluating their own weight of evidence. Morris (1971, 1974, 1977) provided a general framework for recipients to update their own uncertainties after learning an expert's opinion. Although this framework is normative under Bayesian axioms and several forensic scholars advocate the use of Bayesian reasoning, few resources describe its application in forensic science. This paper addresses this gap by examining how recipients can combine principles of science and Bayesian reasoning to evaluate their own likelihood ratios for expert opinions. This exercise helps clarify how an expert's role depends on whether one envisions recipients to be logical and scientific or deferential. Illustrative examples with an expert's opinion expressed as a categorical conclusion, likelihood ratio, or range of likelihood ratios, or with likelihood ratios from multiple experts, each reveal the importance and influence of validation data for logical recipients' interpretations.

stat.AP↗

arXiv2022

A New String Edit Distance and Applications

String edit distances have been used for decades in applications ranging from spelling correction and web search suggestions to DNA analysis. Most string edit distances are variations of the Levenshtein distance and consider only single-character edits. In forensic applications polymorphic genetic markers such as short tandem repeats (STRs) are used. At these repetitive motifs the DNA copying errors consist of more than just single base differences. More often the phenomenon of ``stutter'' is observed, where the number of repeated units differs (by whole units) from the template. To adapt the Levenshtein distance to be suitable for forensic applications where DNA sequence similarity is of interest, a generalized string edit distance is defined that accommodates the addition or deletion of whole motifs in addition to single-nucleotide edits. A dynamic programming implementation is developed for computing this distance between sequences. The novelty of this algorithm is in handling the complex interactions that arise between multiple- and single-character edits. Forensic examples illustrate the purpose and use of the Restricted Forensic Levenshtein (RFL) distance measure, but applications extend to sequence alignment and string similarity in other biological areas, as well as dynamic programming algorithms more broadly.

q-bio.GN↗

arXiv2022

Bayesian Reasoning and Evidence Communication

Many resources for forensic scholars and practitioners, such as journal articles, guidance documents, and textbooks, address how to make a value of evidence assessment in the form of a likelihood ratio (LR) when deciding between two competing propositions. These texts often describe experts presenting their LR values to other parties in the judicial system, such as lawyers, judges, and potentially jurors, but few texts explicitly address how a recipient is expected to utilize the provided LR value. Those that do often imply, or directly suggest, a hybrid modification of Bayes' rule in which a decision maker multiplies their prior odds with another person's assessment of LR to obtain their posterior odds. In this paper, we illustrate how someone adhering to Bayesian reasoning would update their personal uncertainty in response to someone else presenting a personal LR value (or any other form of an opinion) and emphasize that the hybrid approach is a departure from Bayesian reasoning. We further consider implications of recipients adhering to Bayesian reasoning on the role and ideal content of expert's reports and testimony and address published responses to our 2017 paper (Lund and Iyer, 2017), where we previously argued that the hybrid equation is not supported by Bayesian reasoning.

stat.AP↗

arXiv2016

Likelihood Ratio as Weight of Forensic Evidence: A Metrological Perspective

In this article we provide a rebuttal against the possible perception that a single number, such as the Likelihood Ratio, can provide an objective, authoritative or definitive weight of evidence. We also illustrate the extent to which conclusions can vary depending on the assumptions used in the analysis, even under alternative assumptions that are judged to be consistent with available empirical information. To facilitate these goals, we introduce the notion of a Lattice of Assumptions and an Uncertainty Pyramid illustrated in the context of a previously published example involving glass evidence. We take the position that rather than focusing on a single number summary as the weight of evidence it is the duty of the forensic expert to assist the trier of fact in forming their own interpretations from a clear understanding of the objective and demonstrably available information. We hope the presented arguments will inspire those in the forensic science community to pursue establishing their practice on a solid foundation of measurement science.

stat.AP↗