Yajing Luo

Hanqing Wang, Yajing Luo, Boya Xiong et al.

Stylistic headline generation is the task to generate a headline that not only summarizes the content of an article, but also reflects a desired style that attracts users. As style-specific article-headline pairs are scarce, previous researches focus on unsupervised approaches with a standard headline generation dataset and mono-style corpora. In this work, we follow this line and propose StyleBART, an unsupervised approach for stylistic headline generation. Our method decorates the pretrained BART model with adapters that are responsible for different styles and allows the generation of headlines with diverse styles by simply switching the adapters. Different from previous works, StyleBART separates the task of style learning and headline generation, making it possible to freely combine the base model and the style adapters during inference. We further propose an inverse paraphrasing task to enhance the style adapters. Extensive automatic and human evaluations show that StyleBART achieves new state-of-the-art performance in the unsupervised stylistic headline generation task, producing high-quality headlines with the desired style.

Yutao Hou, Yajing Luo, Zhiwen Ruan et al.

Large language models (LLMs) demonstrate remarkable performance across various tasks, prompting researchers to develop diverse evaluation benchmarks. However, existing benchmarks typically measure the ability of LLMs to respond to individual questions, neglecting the complex interactions in real-world applications. In this paper, we introduce Compound Question Synthesis (CQ-Syn) to create the Compound-QA benchmark, focusing on compound questions with multiple sub-questions. This benchmark is derived from existing QA datasets, annotated with proprietary LLMs and verified by humans for accuracy. It encompasses five categories: Factual-Statement, Cause-and-Effect, Hypothetical-Analysis, Comparison-and-Selection, and Evaluation-and-Suggestion. It evaluates the LLM capability in terms of three dimensions including understanding, reasoning, and knowledge. Our assessment of eight open-source LLMs using Compound-QA reveals distinct patterns in their responses to compound questions, which are significantly poorer than those to non-compound questions. Additionally, we investigate various methods to enhance LLMs performance on compound questions. The results indicate that these approaches significantly improve the models' comprehension and reasoning abilities on compound questions.

Yuan Gao, Yajing Luo, Junhong Wang et al.

Humans can easily deduce the relative pose of a previously unseen object, without labeling or training, given only a single query-reference image pair. This is arguably achieved by incorporating i) 3D/2.5D shape perception from a single image, ii) render-and-compare simulation, and iii) rich semantic cue awareness to furnish (coarse) reference-query correspondence. Motivated by this, we propose a novel 3D generalizable relative pose estimation method by elaborating 3D/2.5D shape perception with a 2.5D shape from an RGB-D reference, fulfilling the render-and-compare paradigm with an off-the-shelf differentiable renderer, and leveraging the semantic cues from a pretrained model like DINOv2. Specifically, our differentiable renderer takes the 2.5D rotatable mesh textured by the RGB and the semantic maps (obtained by DINOv2 from the RGB input), then renders new RGB and semantic maps (with back-surface culling) under a novel rotated view. The refinement loss comes from comparing the rendered RGB and semantic maps with the query ones, back-propagating the gradients through the differentiable renderer to refine the 3D relative pose. As a result, \emph{our method can be readily applied to unseen objects, given only a single RGB-D reference, without labeling or training}. Extensive experiments on LineMOD, LM-O, and YCB-V show that our training-free method significantly outperforms the state-of-the-art supervised methods, especially under the rigorous \texttt{Acc@5/10/15}$^\circ$ metrics and the challenging cross-dataset settings.

Yajing Luo, Peng Liang, Chong Wang et al.

Background: In recent years, Low-code development (LCD) is growing rapidly, and Gartner and Forrester have predicted that the use of LCD is very promising. Giant companies, such as Microsoft, Mendix, and Outsystems have also launched their LCD platforms. Aim: In this work, we explored two popular online developer communities, Stack Overflow (SO) and Reddit, to provide insights on the characteristics and challenges of LCD from a practitioners' perspective. Method: We used two LCD related terms to search the relevant posts in SO and extracted 73 posts. Meanwhile, we explored three LCD related subreddits from Reddit and collected 228 posts. We extracted data from these posts and applied the Constant Comparison method to analyze the descriptions, benefits, and limitations and challenges of LCD. For platforms and programming languages used in LCD, implementation units in LCD, supporting technologies of LCD, types of applications developed by LCD, and domains that use LCD, we used descriptive statistics to analyze and present the results. Results: Our findings show that: (1) LCD may provide a graphical user interface for users to drag and drop with little or even no code; (2) the equipment of out-of-the-box units (e.g., APIs and components) in LCD platforms makes them easy to learn and use as well as speeds up the development; (3) LCD is particularly favored in the domains that have the need for automated processes and workflows; and (4) practitioners have conflicting views on the advantages and disadvantages of LCD. Conclusions: Our findings suggest that researchers should clearly define the terms when they refer to LCD, and developers should consider whether the characteristics of LCD are appropriate for their projects.

Xiaofeng Han, Amjed Tahir, Peng Liang et al.

Code review plays an important role in software quality control. A typical review process would involve a careful check of a piece of code in an attempt to find defects and other quality issues/violations. One type of issues that may impact the quality of the software is code smells - i.e., bad programming practices that may lead to defects or maintenance issues. Yet, little is known about the extent to which code smells are identified during code reviews. To investigate the concept behind code smells identified in code reviews and what actions reviewers suggest and developers take in response to the identified smells, we conducted an empirical study of code smells in code reviews using the two most active OpenStack projects (Nova and Neutron). We manually checked 19,146 review comments obtained by keywords search and random selection, and got 1,190 smell-related reviews to study the causes of code smells and actions taken against the identified smells. Our analysis found that 1) code smells were not commonly identified in code reviews, 2) smells were usually caused by violation of coding conventions, 3) reviewers usually provided constructive feedback, including fixing (refactoring) recommendations to help developers remove smells, and 4) developers generally followed those recommendations and actioned the changes. Our results suggest that 1) developers should closely follow coding conventions in their projects to avoid introducing code smells, and 2) review-based detection of code smells is perceived to be a trustworthy approach by developers, mainly because reviews are context-sensitive (as reviewers are more aware of the context of the code given that they are part of the project's development team).