World Robot Challenge 2020 -- Partner Robot: A Data-Driven Approach for Room Tidying with Mobile Manipulator

Matsushima, Tatsuya; Noguchi, Yuki; Arima, Jumpei; Aoki, Toshiki; Okita, Yuki; Ikeda, Yuya; Ishimoto, Koki; Taniguchi, Shohei; Yamashita, Yoichi; Seto, Shoichi; Gu, Shixiang; Iwasawa, Yusuke; Matsuo, Yutaka

doi:10.48550/arxiv.2207.10106

Cited by 2 publications

(2 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Although color and shape recognition is uncommon, class recognition is a common problem. In [16], class recognition using ITR is used in a robot competition because it is possible to improve the recognition accuracy subsequently without re-training by just tuning the text of the choices (called prompt tuning).…”

Section: A Object Recognitionmentioning

confidence: 99%

Learning-Based Wiping Behavior of Low-Rigidity Robots Considering Various Surface Materials and Task Definitions

Kawaharazuka

Kanazawa

Okada

et al. 2022

2022 IEEE-RAS 21st International Conference on Humanoid Robots (Humanoids)

View full text Add to dashboard Cite

In recent years, a number of models that learn the relations between vision and language from large datasets have been released. These models perform a variety of tasks, such as answering questions about images, retrieving sentences that best correspond to images, and finding regions in images that correspond to phrases. Although there are some examples, the connection between these pre-trained vision-language models and robotics is still weak. If they are directly connected to robot motions, they lose their versatility due to the embodiment of the robot and the difficulty of data collection, and become inapplicable to a wide range of bodies and situations. Therefore, in this study, we categorize and summarize the methods to utilize the pre-trained vision-language models flexibly and easily in a way that the robot can understand, without directly connecting them to robot motions. We discuss how to use these models for robot motion selection and motion planning without re-training the models. We consider five types of methods to extract information understandable for robots, and show the results of state recognition, object recognition, affordance recognition, relation recognition, and anomaly detection based on the combination of these five methods. We expect that this study will add flexibility and ease-of-use, as well as new applications, to the recognition behavior of existing robots.

show abstract

Section: A Object Recognitionmentioning

confidence: 99%

Learning-Based Wiping Behavior of Low-Rigidity Robots Considering Various Surface Materials and Task Definitions

Kawaharazuka

Kanazawa

Okada

et al. 2022

2022 IEEE-RAS 21st International Conference on Humanoid Robots (Humanoids)

View full text Add to dashboard Cite

show abstract

“…These robots can smoothly navigate narrow indoor spaces, use complex human voice commands to interact with humans, independently detect objects, and accomplish various home tasks. In [5], the use of HSR applications, through visual recognition, object manipulation, and motion planning technology, was described to achieve home sorting tasks.…”

Section: Introductionmentioning

confidence: 99%

Adaptive Fuzzy Inference Decision Strategy for Service Robots Tidying Up Objects

Liu

et al. 2023

IEEE Access

View full text Add to dashboard Cite

Tidying up objects poses a complex challenge for service robots, particularly when it involves scheduling operations that require real-time action for achieving optimal results. This paper presents a decision system aimed at selecting priority locations to address the problem of robots searching for miscellaneous items in indoor environments. Initially, three datasets with varying complexities are generated using real environmental data, and the environmental information is integrated into feature matrices. Subsequently, the rule base of the fuzzy inference system is trained and optimized using an equilibrium optimizer, and its performance is compared with other commonly used algorithms. The feature matrices and the optimized rule base are then incorporated into the fuzzy inference system, leveraging the traveling salesman problem to determine the optimal sequence for visiting locations. The accuracy and efficiency of the proposed method are validated through physical implementation tests using analog data. Finally, the system is applied and evaluated in real-world scenarios to validate its effectiveness. The video of experiment is available at https://youtu.be/rjMDgopM-9M.INDEX TERMS Autonomous mobile robot, equilibrium optimizer, fuzzy inference system, tidying up object problem, travelling salesman problem.

show abstract

World Robot Challenge 2020 -- Partner Robot: A Data-Driven Approach for Room Tidying with Mobile Manipulator

Cited by 2 publications

References 30 publications

Learning-Based Wiping Behavior of Low-Rigidity Robots Considering Various Surface Materials and Task Definitions

Learning-Based Wiping Behavior of Low-Rigidity Robots Considering Various Surface Materials and Task Definitions

Adaptive Fuzzy Inference Decision Strategy for Service Robots Tidying Up Objects

Contact Info

Product

Resources

About