In our increasingly digital world, there’s a growing demand for 3D models in various fields. Manual 3D model creation is time-consuming and prone to errors, highlighting the need for automation. In this work, we propose a 3D texture transfer framework, 3DTextureNet, to transfer 3D texture from style to content 3D objects, enabling the generation of a wide range of stylized 3D models. We analyze the effects of multiple model hyperparameters on 3D texture transfer. To evaluate the proposed 3D texture transfer framework, we conduct a user study with 3D designers. Our evaluation results demonstrate that our approach effectively transfers 3D texture from style to content objects and the stylized outputs aid in designers’ creativity.

Abstract Lately, crowdsourcing has emerged as a viable option for getting work done by leveraging the collective intelligence of the crowd. With many tasks posted every day, the size of crowdsourcing platforms is growing exponentially. Hence, workers face an important challenge in selecting the right task. Despite the task filtering criteria available on the platform to select the right task, crowd workers find it difficult to choose the most relevant task and must glean through the filtered tasks to find the relevant tasks. In this paper, we propose a framework for recommending tasks to workers. The proposed framework evaluates the worker's fitment over the tasks based on the worker's preference, past tasks he/she has performed, and tasks done by similar workers. We also proposed an approach to estimate the right price for a crowdsourced task for a specific worker. We evaluated our approach on the datasets collected from popular crowdsourcing platforms. Our experimental results show that the recommendation made by our framework for task and price is significantly better as compared with the baseline approach.

Lately, crowdsourcing has emerged as a viable option of getting work done by leveraging the collective intelligence of the crowd. With many tasks posted every day, the size of crowdsourcing platforms is growing exponentially. Hence, workers face an important challenge of selecting the right task. Despite the task filtering criteria available on the platform to select the right task, crowd workers find it difficult to choose the most relevant task and must glean through the filtered tasks to find the relevant tasks. In this paper, we propose a framework for recommending tasks to workers. The proposed framework evaluates the worker's fitment over the tasks based on worker's preference, past tasks (s)he has performed, and tasks done by similar workers. We evaluated our approach on the datasets collected from popular crowdsourcing platform. Our experimental results based on 5,000 tasks and 3,000 workers show that the recommendation made by our framework is significantly better as compared to the baseline approach.

Crowdsourced testing is gaining a lot of attention these days. Crowdsourced testing utilizes testers which subscribe to an external or internal crowdsourcing platform. Usually these testers are distributed across geographies. Thus, such testing can be treated as a form of distributed testing. Crowdsourced testing, quite often, is used to perform exploratory testing in which testers test the features as per their wish. Such testing results in redundant testing efforts due to the lack of awareness of other testers' activities, thus does not yield the full benefits of crowdsourced testing in terms of speed and coverage. Moreover, as each tester possesses a testing style, the present model of crowdsourced testing does not fully utilize tester's strength or style. In this paper, we study various redundancies involved in a distributed testing. We also study testers' behavior to understand various testing styles. The study finds that there exists a lot of redundancies in testing efforts. The study also observes that testers indeed have different testing styles which should be understood more deeply for engaging them better. Our study shows that there is need for tools and strategies for allocating testing tasks that leverages testers' testing style and creates more awareness among the testers.

The emergence of online labor markets has concentrated a lot of attention on the prospect of using crowdsourcing for software development, with a potential to reduce costs, improve time-to-market, and access high-quality skills on demand. However, crowdsourcing of software development is still not widely adopted. A key barrier to adoption is a lack of confidence that a task will be completed on time with the required quality standards. While good managers can develop good, intuitive estimates of task completion when assigning work to their team members, they might lack similar intuition for individuals drawn from an online crowd. The phrase, "Post and Hope" is thus sometimes used when talking about the crowdsourcing of software-development tasks. The objective of this paper is to show the value of replacing the traditional, intuitive assessment of a team's capability with a quantitative assessment of the crowd, derived through analysis of historical performance on similar tasks. This analysis will serve to transform "Post and Hope" to "Post and Expect." We demonstrate this by analyzing data about tasks performed on two popular crowdsourcing platforms: Topcoder and Upwork. Analysis of historical data from these platforms indicates that the platforms indeed demonstrate some level of predictability in task completion. We have identified certain factors that consistently contribute to task completion on both the platforms. Our findings suggest that a data-driven decision processes can play an important role in successful adoption of crowdsourcing practice for software development.

This paper proposes an approach for adopting alternative workforce in an organization. Alternative workforce refers to a pool of workers who work for the organization as contract workers or as crowd workers for a set of specific tasks or duration. Adoption of crowd workers as an alternative workforce is gaining a lot of attention these days. However, it is still not widely adopted by big organizations because of the concerns related to quality, timeliness, and confidentiality. A partial adoption of crowd workforce is a natural next step to leverage the benefits of crowdsourcing. The above partial adoption creates a hybrid workforce structure where different type of workers, such as full time employees, contractors, and crowd workers, work for the organization. A number of challenges need to be addressed for the above model to succeed. For instance, hiring right workers, establishing a proper collaboration among the workers distributed across geographies, and assessing the workers for confidentiality and privacy. This paper proposes a platform that alleviates some of the above challenges. We present a pilot performed on the platform and initial experiences gained from the adoption of the platform.

Adoption of software engineering best practices requires a good infrastructure for monitoring compliance with each practice. Quite often, practice compliance implies that a set of policies are being adhered by each development group. Monitoring each practice to the level of individual policies is cumbersome as it involves some one to follow up manually with the development groups on each policy in each practice. This results in a significant quality assurance effort in a big organization where we have large number of development groups. Moreover, the data can be error prone and outdated due to time and multiple people involved in such activities. In this paper, we propose a policy checking and reporting framework to automate the reporting to a good extent. The framework is based on a central reporting service which collects software engineering data from various sources and provides the assessment on each policy related to static analysis practice.

Global nature of multinational organizations pose a different set of challenges in an effective adoption of software engineering tools. Software projects are spread across multiple geographies and tools are deployed globally in these organizations. Global tool deployment is often motivated by an attractive enterprise cost of the tool. Moreover, it helps in harmonizing practices across development groups and useful in establishing benchmarks without further normalizations. There are various parameters to drive a successful tool deployment in a global setup. For example, a good roll out plan, knowledge sharing platform, coordination and trust building among the change agents from different geographies, and a good central support. In this paper, we present our experience of a global roll out of a software engineering tool in a large organization. It has been observed that besides tool's features, factors such as knowledge sharing, active feedback, and training plays an important role in an effective tool deployment.

In order to maximize cost and quality gains, industrial automation systems need to incorporate the best practices of software engineering in their application development process. However, this requires the right set of tools and methodologies that cater to the needs of the automation domain. While there have been a few efforts towards applying state-of-the-art software engineering tools and techniques to the automation domain, these have not been universally adopted. This paper discusses some of the challenges in adopting software engineering principles for industrial automation application development. Further, the paper presents a case for research activities to look for more practical solutions for industrial applications.

Industrial automation applications are often reengineered to serve purposes such as reducing the load on controllers by adding additional controllers in the system, improving the throughput / performance of the application, or migrating to a new platform. We present an approach for automatically computing reengineering options for a legacy industrial automation application over an additional set of controllers. An engineer, while reengineering an application over an additional set of controllers, needs to compute the various dependencies between the modules (such as Program organization units, function blocks, statements) of the application. Additionally, he needs to consider aspects, such as, communication overhead, controller load, and geographical placement of field (input/output) devices. This is a time consuming activity when performed manually. Moreover, it is difficult to compare multiple reengineering options, on the above aspects, without some tool support. We propose an approach that automatically computes multiple reengineering options and recommends them to the engineer.

We propose a technique for summarizing Web reviews. Information summarization has become an important problem in the current content saturated world. One such example is the World Wide Web which provides a platform to publish and evaluate information. This collaborative nature of the Web has enabled users to write their opinion on certain topics and also evaluate others' opinions by assigning ranks. In this paper we show that the above aspect of Web can be utilized to generate more useful summary. We consider the problem of generating summary from the Web reviews and the rank (usefulness) assigned to these reviews by other users. We study the usefulness of user ranks in the summarization task. Based on the study, we propose a technique which takes ranked reviews as input and generates a summary. We experiment with different variations of the proposed technique and evaluate them based on different criteria.