arXiv subjects

Jiaxing Lu

Publications and source records attributed to Jiaxing Lu.

6 recordsLinked to original sources

arXiv2025

Approximate Gaussian Mapping for Generative Image Steganography

Ordinary differential equation (ODE)-based diffusion models enable deterministic image synthesis, establishing a reversible mapping suitable for generative steganography. While prevailing methods strictly adhere to a standard normal prior, empirical evidence indicates that controlled deviations from this distribution reduce numerical inversion errors without compromising perceptual quality. Leveraging this observation, the Approximate Gaussian Mapping (AGM) is proposed as a linear transformation strategy that embeds secrets by modulating noise scale and variance. To balance retrieval numerical consistence and security, a two-stage decoupled optimization strategy is introduced to minimize the Kullback-Leibler divergence subject to target bit accuracy constraints. Beyond the proposed method, we conduct a mechanistic analysis of the divergent behaviors between pixel-space and latent-space architectures. The experimental results reveal that the VAE encoder enhances robustness by filtering external perturbations, whereas the structural regularization of the VAE decoder and the semantic variance introduced by text prompts jointly mask embedding artifacts to improve security. Experimental results confirm that pixel-space mplementations maximize embedding capacity for lossless channels, while latent-space approaches offer superior robustness and security suitable for adversarial environments

cs.CR↗

arXiv2024

Constructing Mechanical Design Agent Based on Large Language Models

Since ancient times, mechanical design aids have been developed to assist human users, aimed at improving the efficiency and effectiveness of design. However, even with the widespread use of contemporary Computer-Aided Design (CAD) systems, there are still high learning costs, repetitive work, and other challenges. In recent years, the rise of Large Language Models (LLMs) has introduced new productivity opportunities to the field of mechanical design. Yet, it remains unrealistic to rely on LLMs alone to complete mechanical design tasks directly. Through a series of explorations, we propose a method for constructing a comprehensive Mechanical Design Agent (MDA) by guiding LLM learning. To verify the validity of our proposed method, we conducted a series of experiments and presented relevant cases.

cs.CE↗

arXiv2023

Let's Discover More API Relations: A Large Language Model-based AI Chain for Unsupervised API Relation Inference

APIs have intricate relations that can be described in text and represented as knowledge graphs to aid software engineering tasks. Existing relation extraction methods have limitations, such as limited API text corpus and affected by the characteristics of the input text.To address these limitations, we propose utilizing large language models (LLMs) (e.g., GPT-3.5) as a neural knowledge base for API relation inference. This approach leverages the entire Web used to pre-train LLMs as a knowledge base and is insensitive to the context and complexity of input texts. To ensure accurate inference, we design our analytic flow as an AI Chain with three AI modules: API FQN Parser, API Knowledge Extractor, and API Relation Decider. The accuracy of the API FQN parser and API Relation Decider module are 0.81 and 0.83, respectively. Using the generative capacity of the LLM and our approach's inference capability, we achieve an average F1 value of 0.76 under the three datasets, significantly higher than the state-of-the-art method's average F1 value of 0.40. Compared to CoT-based method, our AI Chain design improves the inference reliability by 67%, and the AI-crowd-intelligence strategy enhances the robustness of our approach by 26%.

cs.SE↗

arXiv2022

SE Factual Knowledge in Frozen Giant Code Model: A Study on FQN and its Retrieval

Pre-trained giant code models (PCMs) start coming into the developers' daily practices. Understanding what types of and how much software knowledge is packed into PCMs is the foundation for incorporating PCMs into software engineering (SE) tasks and fully releasing their potential. In this work, we conduct the first systematic study on the SE factual knowledge in the state-of-the-art PCM CoPilot, focusing on APIs' Fully Qualified Names (FQNs), the fundamental knowledge for effective code analysis, search and reuse. Driven by FQNs' data distribution properties, we design a novel lightweight in-context learning on Copilot for FQN inference, which does not require code compilation as traditional methods or gradient update by recent FQN prompt-tuning. We systematically experiment with five in-context-learning design factors to identify the best in-context learning configuration that developers can adopt in practice. With this best configuration, we investigate the effects of amount of example prompts and FQN data properties on Copilot's FQN inference capability. Our results confirm that Copilot stores diverse FQN knowledge and can be applied for the FQN inference due to its high inference accuracy and non-reliance on code analysis. Based on our experience interacting with Copilot, we discuss various opportunities to improve human-CoPilot interaction in the FQN inference task.

cs.SE↗

arXiv2022

STAR: Zero-Shot Chinese Character Recognition with Stroke- and Radical-Level Decompositions

Zero-shot Chinese character recognition has attracted rising attention in recent years. Existing methods for this problem are mainly based on either certain low-level stroke-based decomposition or medium-level radical-based decomposition. Considering that the stroke- and radical-level decompositions can provide different levels of information, we propose an effective zero-shot Chinese character recognition method by combining them. The proposed method consists of a training stage and an inference stage. In the training stage, we adopt two similar encoder-decoder models to yield the estimates of stroke and radical encodings, which together with the true encodings are then used to formalize the associated stroke and radical losses for training. A similarity loss is introduced to regularize stroke and radical encoders to yield features of the same characters with high correlation. In the inference stage, two key modules, i.e., the stroke screening module (SSM) and feature matching module (FMM) are introduced to tackle the deterministic and confusing cases respectively. In particular, we introduce an effective stroke rectification scheme in FMM to enlarge the candidate set of characters for final inference. Numerous experiments over three benchmark datasets covering the handwritten, printed artistic and street view scenarios are conducted to demonstrate the effectiveness of the proposed method. Numerical results show that the proposed method outperforms the state-of-the-art methods in both character and radical zero-shot settings, and maintains competitive performance in the traditional seen character setting.

cs.CV↗

arXiv2019

HDI-Forest: Highest Density Interval Regression Forest

By seeking the narrowest prediction intervals (PIs) that satisfy the specified coverage probability requirements, the recently proposed quality-based PI learning principle can extract high-quality PIs that better summarize the predictive certainty in regression tasks, and has been widely applied to solve many practical problems. Currently, the state-of-the-art quality-based PI estimation methods are based on deep neural networks or linear models. In this paper, we propose Highest Density Interval Regression Forest (HDI-Forest), a novel quality-based PI estimation method that is instead based on Random Forest. HDI-Forest does not require additional model training, and directly reuses the trees learned in a standard Random Forest model. By utilizing the special properties of Random Forest, HDI-Forest could efficiently and more directly optimize the PI quality metrics. Extensive experiments on benchmark datasets show that HDI-Forest significantly outperforms previous approaches, reducing the average PI width by over 20% while achieving the same or better coverage probability

cs.LG↗