arXiv subjects

Summit Haque

Publications and source records attributed to Summit Haque.

6 recordsLinked to original sources

arXiv2026

Handcrafted Feature Fusion for Reliable Detection of AI-Generated Images

The rapid progress of generative models has enabled the creation of highly realistic synthetic images, raising concerns about authenticity and trust in digital media. Detecting such fake content reliably is an urgent challenge. While deep learning approaches dominate current literature, handcrafted features remain attractive for their interpretability, efficiency, and generalizability. In this paper, we conduct a systematic evaluation of handcrafted descriptors, including raw pixels, color histograms, Discrete Cosine Transform (DCT), Histogram of Oriented Gradients (HOG), Local Binary Patterns (LBP), Gray-Level Co-occurrence Matrix (GLCM), and wavelet features, on the CIFAKE dataset of real versus synthetic images. Using 50,000 training and 10,000 test samples, we benchmark seven classifiers ranging from Logistic Regression to advanced gradient-boosted ensembles (LightGBM, XGBoost, CatBoost). Results demonstrate that LightGBM consistently outperforms alternatives, achieving PR-AUC 0.9879, ROC-AUC 0.9878, F1 0.9447, and a Brier score of 0.0414 with mixed features, representing strong gains in calibration and discrimination over simpler descriptors. Across three configurations (baseline, advanced, mixed), performance improves monotonically, confirming that combining diverse handcrafted features yields substantial benefit. These findings highlight the continued relevance of carefully engineered features and ensemble learning for detecting synthetic images, particularly in contexts where interpretability and computational efficiency are critical.

cs.CV↗

arXiv2026

Whitespaces Don't Lie: Feature-Driven and Embedding-Based Approaches for Detecting Machine-Generated Code

Large language models (LLMs) have made it remarkably easy to synthesize plausible source code from natural language prompts. While this accelerates software development and supports learning, it also raises new risks for academic integrity, authorship attribution, and responsible AI use. This paper investigates the problem of distinguishing human-written from machine-generated code by comparing two complementary approaches: feature-based detectors built from lightweight, interpretable stylometric and structural properties of code, and embedding-based detectors leveraging pretrained code encoders. Using a recent large-scale benchmark dataset of 600k human-written and AI-generated code samples, we find that feature-based models achieve strong performance (ROC-AUC 0.995, PR-AUC 0.995, F1 0.971), while embedding-based models with CodeBERT embeddings are also very competitive (ROC-AUC 0.994, PR-AUC 0.994, F1 0.965). Analysis shows that features tied to indentation and whitespace provide particularly discriminative cues, whereas embeddings capture deeper semantic patterns and yield slightly higher precision. These findings underscore the trade-offs between interpretability and generalization, offering practical guidance for deploying robust code-origin detection in academic and industrial contexts.

cs.SE↗

arXiv2025

Code Comprehension with GitHub Copilot: Performance Gains, Comprehension Trade-offs, and Behavioral Predictors in Brownfield Programming

Teaching Computer Science (CS) students how to comprehend and maintain legacy code bases is a critical challenge in software engineering education. While Generative AI (GenAI) assistants like GitHub Copilot improve task completion speed and correctness, their impact on code understanding remains unclear. We conducted a within-subject study with 15 graduate CS students completing feature implementation tasks with and without Copilot. Despite significant performance improvements, participants showed no overall comprehension improvement ($p=0.59$), revealing a \textit{comprehension-performance decoupling}. Further analysis uncovered a \textit{comprehension trade-off}: performance gains negatively correlated with reverse engineering comprehension ($\rho=-0.57$, $p=0.026$) but showed a positive trend with implementation comprehension ($\rho=0.50$, $p=0.06$). A follow-up behavioral analysis revealed that \textit{how} students used Copilot determined outcomes: Engaging in verification loops in which programmers actively reviewed generated code strongly predicted comprehension ($p<0.001$, $r=0.96$), with high-comprehension participants verifying code 4.7 times more frequently than low-comprehension participants. These findings suggest that GenAI tools do not inherently undermine comprehension; rather, passive consumption patterns do. This suggests a need to alter programming education to teach system-level verification skills, and the need to redesign educational GenAI tools to scaffold active cognitive engagement.

cs.SE↗

arXiv2025

The Effects of GitHub Copilot on Computing Students' Programming Effectiveness, Efficiency, and Processes in Brownfield Programming Tasks

When graduates of computing degree programs enter the software industry, they will most likely join teams working on legacy code bases developed by people other than themselves. In these so-called brownfield software development settings, generative artificial intelligence (GenAI) coding assistants like GitHub Copilot are rapidly transforming software development practices, yet the impact of GenAI on student programmers performing brownfield development tasks remains underexplored. This paper investigates how GitHub Copilot influences undergraduate students' programming performance, behaviors, and understanding when completing brownfield programming tasks in which they add new code to an unfamiliar code base. We conducted a controlled experiment in which 10 undergraduate computer science students completed highly similar brownfield development tasks with and without Copilot in a legacy web application. Using a mixed-methods approach combining performance analysis, behavioral analysis, and exit interviews, we found that students completed tasks 35% faster (p < 0.05) and made 50% more solution progress p (< 0.05) when using Copilot. Moreover, our analysis revealed that, when using Copilot, students spent 11% less time manually writing code (p < 0.05), and 12% less time conducting web searches (p < 0.05), providing evidence of a fundamental shift in how they engaged in programming. In exit interviews, students reported concerns about not understanding how or why Copilot suggestions work. This research suggests the need for computing educators to develop new pedagogical approaches that leverage GenAI assistants' benefits while fostering reflection on how and why GenAI suggestions address brownfield programming tasks. Complete study results and analysis are presented at https://ghcopilot-icer.github.io/.

cs.SE↗

arXiv2023

Multi-Agent Join

It is crucial to provide real-time performance in many applications, such as interactive and exploratory data analysis. In these settings, users often need to view subsets of query results quickly. It is challenging to deliver such results over large datasets for relational operators over multiple relations, such as join. Join algorithms usually spend a long time on scanning and attempting to join parts of relations that may not generate any result. Current solutions usually require lengthy and repeated preprocessing, which is costly and may not be possible to do in many settings. Also, they often support restricted types of joins. In this paper, we outline a novel approach for achieving efficient join processing in which a scan operator of the join learns during query execution, the portions of its relations that might satisfy the join predicate. We further improve this method using an algorithm in which both scan operators collaboratively learn an efficient join execution strategy. We also show that this approach generalizes traditional and non-learning methods for joining. Our extensive empirical studies using standard benchmarks indicate that this approach outperforms similar methods considerably.

cs.DB↗

arXiv2021

Fundamental Machine Learning Routines as Quantum Algorithms on a Superconducting Quantum Computer

The Harrow-Hassidim-Lloyd algorithm is intended for solving the system of linear equations on quantum devices. The exponential advantage of the algorithm comes with four caveats. We present a numerical study of the performance of the algorithm when these caveats are not perfectly matched. We observe that, between diagonal and non-diagonal matrices, the algorithm performs with higher success probability for the diagonal matrices. At the same time, it fails to perform well on lower or higher density sparse Hermitian matrices. Again, Quantum Support Vector Machine algorithm is a promising algorithm for classification problem. We have found out that it works better with binary classification problem than multi-label classification problem. And there are many opportunities left for improving the performance.

quant-ph↗