Abstract

Abstract This document contains the details of the authors' submission to the proceedings of SemEval 2024's Task 8: Multigenerator, Multidomain, and Multilingual Black-Box Machine-Generated Text Detection Subtask A (monolingual) and B. Detection of machine-generated text is becoming an increasingly important task, with the advent of large language models (LLMs). In this paper, we lay out how using weighted averages of RoBERTa layers lets us capture information about text that is relevant to machine-generated text detection

Abstract

We review the main applications of machine learning models that are not fully supervised in particle physics, i.e., clustering, anomaly detection, detector simulation, and unfolding. Unsupervised methods are ideal for anomaly detection tasks—machine learning models can be trained on background data to identify deviations if we model the background data precisely. The learning can also be partially unsupervised when we can provide some information about the anomalies at the data level. Generative models are useful in speeding up detector simulations—they can mimic the computationally intensive task without large resources. They can also efficiently map detector-level data to parton-level data (i.e., data unfolding). In this review, we focus on interesting ideas and connections and briefly overview the underlying techniques wherever necessary.

Abstract

Abstract—Safety-critical real-time systems are conventionally designed to meet specific hard real-time deadline requirements. Many applications in such systems consist of task chains, which exhibit complex temporal dependencies based on the activation rates of the tasks. As a result, there has been an increasing interest in analyzing the age of data (i.e., the time elapsed since the data was first updated) that propagates through the task chains as an important system performance parameter. Transient faults in safety-critical systems can lead to undesirable consequences. One of the most common methods used for fault tolerance in such systems is checkpointing. However, the interference due to checkpointing can significantly impact the data age, reducing the predictability of the system’s performance. This paper proposes an analytical framework for calculating the data age in real-time task chains with checkpointing for transient faults in both single-core and multi-core platforms. The proposed analysis has been validated by comparing it with extensive simulations of the execution of real-time task chain- based systems under faults. It shows comparable data age values with low time complexity.

Abstract

Random Numbers are widely employed in cryptography and security applications. This paper presents a novel approach to generate high-quality random bitstreams by harnessing the inherent noise properties of ring oscillators. We implemented ring oscillators with varying numbers of stages (3, 5, and 7), different geometries and different startup voltages in Cadence and recorded their total output power, which includes the cumulative noise effects. Subsequently, we exported these power measurements to MATLAB, where we applied a Fast Fourier Transform (FFT)-based technique to extract the total noise characteristics for each ring oscillator. Using the obtained noise data, we generated separate random bitstreams of 10 million bits for the 3-stage, 5-stage, and 7-stage ring oscillators. The final random bitstream, consisting of 10 million bits, was created by performing a bitwise XOR operation on the bitstreams generated by each ring oscillator. The degree of randomness of the generated bitstreams was assessed using the NIST 800-22 statistical test suite. Remarkably, the final random bitstream exhibited strong robustness and suitability for cryptographic applications. This innovative approach leverages the noise properties of ring oscillators to create reliable random bitstreams, offering potential applications in secure communications and cryptography. The results highlight the feasibility of using ring oscillators as noise sources for random bit generation and underscore their effectiveness in meeting stringent randomness criteria

Abstract

In addition to innate human intelligence, having access to extensive context and world knowledge is a crucial factor that aids in comprehending natural language, making it smooth and effortless to understand words with multiple meanings for humans. Although machines lack intrinsic intelligence, their capacity to learn language can greatly improve with access to more data, which serves as valuable context. In Natural Language Processing (NLP), the task of identifying and attributing the right sense of a word in a given context is called Word Sense Disambiguation (WSD). WSD, as a sub-task, plays a crucial role in several NLP applications such as Machine Translation. Every language has a set of words that have multiple senses. Sanskrit, one of the ancient and classical languages of the Indian subcontinent is no exception to this. Like many other languages with a rich literary tradition, Sanskrit features a multitude of polysemous words. However, it is essential to acknowledge that the data used to train machine models on Sanskrit is considerably less compared to European and a few other Indian languages. Consequently, the task of disambiguating word senses in Sanskrit presents a highly complex challenge for machines, especially when considering the unique and rich nature of its literary language. The purpose of this paper is to delineate the potential areas where the infusion of additional data can enhance language learning, through a manual error analysis taxonomy focused on the Bhagavadgītā. Our analysis will delve into the translation outcomes produced by Google Translate, which is considered the state-of-the-art tool for handling Sanskrit and other languages with limited available resources.

Abstract

Intelligent vehicle systems require a deep understanding of the interplay between road conditions, surrounding entities, and the ego vehicle's driving behavior for safe and efficient navigation. This is particularly critical in developing countries where traffic situations are often dense and unstructured with heterogeneous road occupants. Existing datasets, predominantly geared towards structured and sparse traffic scenarios, fall short of capturing the complexity of driving in such environments. To fill this gap, we present IDD-X, a large-scale dual-view driving video dataset. With 697K bounding boxes, 9K important object tracks, and 1-12 objects per video, IDD-X offers comprehensive ego-relative annotations for multiple important road objects covering 10 categories and 19 explanation label categories. The dataset also incorporates rearview information to provide a more complete representation of the driving environment. We also introduce custom-designed deep networks aimed at multiple important object localization and per-object explanation prediction. Overall, our dataset and introduced prediction models form the foundation for studying how road conditions and surrounding entities affect driving behavior in complex traffic situations.

Abstract

Characters are an important aspect of any storyline and identifying and including them in descriptions is necessary for story understanding. While previous work has largely ignored identity and generated captions with someone (anonymized names), recent work formulates id-aware captioning as a fill-in-the-blanks (FITB) task, where, given a caption with blanks, the goal is to predict person id labels. However, to predict captions with ids, a two-stage approach is required: first predict captions with someone, then fill in identities. In this work, we present a new single stage approach that can seamlessly switch between id-aware caption generation or FITB when given a caption with blanks. Our model, Movie-Identity Captioner (MICap), uses a shared auto-regressive decoder that benefits from training with FITB and full-caption generation objectives, while the encoder can benefit from or disregard captions with blanks as input. Another challenge with id-aware captioning is the lack of a metric to capture subtle differences between person ids. To this end, we introduce iSPICE, a caption evaluation metric that focuses on identity tuples created through intermediate scene graphs. We evaluate MICap on Large-Scale Movie Description Challenge (LSMDC), where we show a 4.2% improvement in FITB accuracy, and a 1-2% bump in classic captioning metrics.

Abstract

The default approach to deal with the enormous size and limited accessibility of many Web and social media networks is to sample one or more subnetworks from a conceptually unbounded unknown network. Clearly, the extracted subnetworks will crucially depend on the sampling scheme. Motivated by studies of homophily and opinion formation, we propose a variant of snowball sampling designed to prioritize the inclusion of entire cohesive communities rather than any kind of representativeness, breadth, or depth of coverage. The method is illustrated on a concrete example, and experiments on synthetic networks suggest that it behaves as desired

Abstract

No Results Found

Abstract

Identifying keywords in an open-vocabulary context is crucial for personalizing interactions with smart devices. Previous approaches to open vocabulary keyword spotting dependon a shared embedding space created by audio and text encoders. However, these approaches suffer from heterogeneous modality representations (i.e., audio-text mismatch). To address this issue, our proposed framework leverages knowledge acquired from a pre-trained text-to-speech (TTS) system. This knowledge transfer allows for the incorporation of awareness of audio projections into the text representations derived from the text encoder. The performance of the proposed approach is compared with various baseline methods across four different datasets. The robustness of our proposed model is evaluated by assessing its performance across different word lengths and in an Out-of-Vocabulary (OOV) scenario. Additionally, the effectiveness of transfer learning from the TTS system is investigated by analyzing its different intermediate representations. The experimental results indicate that, in the challenging LibriPhrase Hard dataset, the proposed approach outperformed the cross-modality correspondence detector (CMCD) method by a significant improvement of 8.22% in area under the curve (AUC) and 12.56% in equal error rate (EER).