| Project/Area Number |
21K14115
|
|---|
| Research Category |
Grant-in-Aid for Early-Career Scientists
|
| Allocation Type | Multi-year Fund |
|---|
| Review Section |
Basic Section 20020:Robotics and intelligent system-related
|
|---|
| Research Institution | Tohoku University |
|---|
Principal Investigator |
Ravankar Ankit 東北大学, 工学研究科, 特任准教授 (40778528)
|
|---|
| Project Period (FY) |
2021-04-01 – 2025-03-31
|
| Project Status |
Granted (Fiscal Year 2023)
|
| Budget Amount *help |
¥4,030,000 (Direct Cost: ¥3,100,000、Indirect Cost: ¥930,000)
Fiscal Year 2023: ¥910,000 (Direct Cost: ¥700,000、Indirect Cost: ¥210,000)
Fiscal Year 2022: ¥780,000 (Direct Cost: ¥600,000、Indirect Cost: ¥180,000)
Fiscal Year 2021: ¥2,340,000 (Direct Cost: ¥1,800,000、Indirect Cost: ¥540,000)
|
|---|
| Keywords | Semantic SLAM / Probability Inferencing / Object Search / Path Planning / Navigation / Service Robot / SLAM / Deep Learning / Semantic Understanding / Robotics / Mobile Robots / Semantic Mapping / Scene Understanding / 自律ロボティクス / 人工知能 / HCI |
|---|
| Outline of Research at the Start |
本研究では「シーン外の経験」を導入することで新しい課題を処理できるセマンティックSLAMシステムを構築することを目的とする。このスキームによってロボットは新しい環境で効率的に学習することが可能となる。本研究は、認知科学の分野で議論されている「セマンティックシーンの理解」と「物体認識」の知見にインスピレーションを得ており、Deep Reinforcement Learning 構造を利用することによって現在のAIに足りない「少ないデータ」での「フレキシブル」な学習の実現を目指す。
|
| Outline of Annual Research Achievements |
In FY2023, we developed object search strategies for service robots in a simulated indoor living lab environment. We designed two distinct methods: one based on heuristic object search and the other on probabilistic inferencing. These techniques are crucial for tasks like object retrieval and detection in home like settings, where users might ask robots to locate specific items. Our approach integrated open-source datasets to train object classifications, incorporating them into our semantic life-long mapping framework.
Additionally, we created room and region classification algorithms to establish semantic connections between objects, places, and regions. This resulted in a three-layer semantic map, enhancing robot navigation and human robot interaction.
|
| Current Status of Research Progress |
Current Status of Research Progress
2: Research has progressed on the whole more than it was originally planned.
Reason
The project is progressing smoothly but further experiments and data collection were required for evaluating the algorithms. This involved long-term data collection in the simulated living lab at Tohoku University. As a result, the project was extended for an year to complete the remaining task and gather the necessary empirical data.
|
| Strategy for Future Research Activity |
This FY the plan is to do a long-term study inside the living lab and collecting large data for robot planning and navigation. The layered semantic navigation framework is already completed, and we plan to use it for testing object detection tasks for a service robot in the living lab. An evaluative comparison of searching using heuristic and by using probabilistic inferencing will be done and compared with base-line search algorithms. The results are planned to be published in two international journals before the end of the this fiscal year.
|
