Part-based 3D model retrieval queried by keywords representing parts
| Project/Area Number |
16K16055
|
|---|
| Research Category |
Grant-in-Aid for Young Scientists (B)
|
| Allocation Type | Multi-year Fund |
|---|
| Research Field |
Multimedia database
|
|---|
| Research Institution | University of Yamanashi |
|---|
Principal Investigator |
|
|---|
| Project Period (FY) |
2016-04-01 – 2019-03-31
|
| Project Status |
Completed (Fiscal Year 2018)
|
| Budget Amount *help |
¥4,030,000 (Direct Cost: ¥3,100,000、Indirect Cost: ¥930,000)
Fiscal Year 2018: ¥650,000 (Direct Cost: ¥500,000、Indirect Cost: ¥150,000)
Fiscal Year 2017: ¥1,430,000 (Direct Cost: ¥1,100,000、Indirect Cost: ¥330,000)
Fiscal Year 2016: ¥1,950,000 (Direct Cost: ¥1,500,000、Indirect Cost: ¥450,000)
|
|---|
| Keywords | 3次元形状類似検索 / 部分検索 / キーワード検索 / 3次元形状解析 / 深層学習 / 3次元モデル検索 / キーワードタグ付与 / 3次元形状特徴量 / 局所特徴量 / 情報検索 / 機械学習 |
|---|
| Outline of Final Research Achievements |
This research addressed the problem of part-based 3D shape retrieval, which searches, from a database, 3D models that include parts a user desires. To achieve practical part-based 3D shape retrieval, I developed a 3D shape retrieval system queried by “keywords representing parts”. In particular, I focused on developing an auto-tagging algorithm for partial 3D shapes. Proposed algorithm associates geometric features of partial 3D shapes and semantic features of keywords by using deep learning technology. Tagging to partial 3D shapes is processed based on distances among the associated features. Empirical evaluation demonstrated that the proposed algorithm yields nearly 70% accuracy for tagging to partial 3D shapes and retrieval queried by keywords. This accuracy would be sufficient for some application scenarios.
|
| Academic Significance and Societal Importance of the Research Achievements |
3D形状類似検索の問題に対し，部分を表すキーワードという新たな問い合わせ手段を提供した点に意義がある．例えば，工業製品の設計において，本技術が機械CADデータの再利用効率を大きく改善すると期待できる．また，3Dプリンタで印刷可能な3Dモデルには「一部に羽の形を持つ指輪」等，部分的な装飾に趣向を凝らしたものが多い．本技術を利用すれば，これらの3Dモデルを効率的に探し出すことができる．さらには，医療診断において，脳血管や内臓の3D画像を蓄積したデータベースから，「動脈瘤」や「腫瘍」等のキーワードで検索できれば，経験の浅い医師の診断補助や学習支援に役立つと期待できる．
|
Report
(4 results)
Research Products
(20 results)
