- Developing algorithms to improve search accuracy. reduce  semantic  distance towards learning distance in image retrieval.
+ Using correlation information through mahalanobis distance learning.
+ Improving CBIR performance WITH Relevance Feedback using SVM.
+ Using local information with co-training image lookup techniques and meaningful spatial image lookups.
- implementing a CBIR program with distance metric learning in the Corel Photo Gallery Experimental results reported in the Corel Photo Gallery (about 30, 000 images).

Theoretical results:
+ Research on image retrieval in the optimal direction.
+ Research on feature extraction method
+ Research on manifold learning
+ Research improving distance learning in image retrieval
+ Research on feedback related to SVM in image retrieval
+ Research improving image retrieval based on the content of using manifold.
+ Research improving image retrieval with co-training model.
+ Research on clustering in image retrieval
+ Implementing a program learning distance metric with an international image set of  31.695 images commonly used by Corel.
+ Support for 02 PhD student and 05 students in the research direction of the topics.
+ Publishing 05 scientific papers related to the research content of the topic. In which: 02 international papers on SCIE list, 02 paper in Scopus list, 01 paper on Proceedings of national conference.
Applied results:
- Implementing a CBIR program with distance metric learning in the Corel Photo Gallery Experimental results reported in the Corel Photo Gallery (about 30, 000 images).

- To propose manifold and semi-supervised learning method for image lookup combining related feedback with SVM.
- Detect false data for cluster synthesis in WSNM Data identification error for cluster in WSN Based on improving SVM classifier capabilities.
- Improve The Efficiency Of Content-based Image Retrieval Through Incremental Clustering
- Improving the efficiency of image-based image searching Improved the efficiency of content-based images through increased clustering,
- Implementing a CBIR program with distance metric learning in the Corel Photo Gallery Experimental results reported in the Corel Photo Gallery (about 30, 000 images)

-    Scientific papers in referred journals (list):
02 papers in the SCIE, 02 papers in the Scopus, 01 paper in the Proceeding of
Fair 2019 conference.
[1]    Huu, Quynh Nguyen;  Viet, Dung Cu,  Thuy, Quynh Dao Thi,  Quoc, Tao  Ngo, Van, Canh Phuong, “Graph-based semisupervised and manifold learning for image retrieval with SVM-based relevant feedback”, Journal of Intelligent and Fuzzy Systems 37(1)  (SCI-E): 711-722 (2019).
[2]    Thi-Kien Dao, Trong-The Nguyen, Jeng-Shyang Pan, Yu Qiao, And Quoc- Anh Lai, “Identification Failure Data for Cluster Heads Aggregation in WSN Based on Improving Classification of SVM”, IEEE Access , Volume: 8. 2020, pp. 61070-61084.
[3]    Quynh Dao Thi Thuy, Quynh Nguyen Huu, Phuong Nguyen Thi Lan, Tao Ngo Quoc and Minh-Huong Ngo, “Improve The Efficiency Of Content-based Image Retrieval Through Incremental Clustering” Journal of Information Hiding and Multimedia Signal Processing, Vol. 11, No. 3, pp. 103-115, September 2020.
[4]    Phuong Nguyen Thi Lan, Tao Ngo Quoc, Quynh Dao Thi Thuy, Minh-Huong Ngo, “Improve the Effectiveness of Image Retrieval by Combining the Optimal Distance and Linear Discriminant Analysis”, (IJACSA) International Journal of Advanced Computer Science and Applications, Vol. 12, No. 2, 2021.
[5]    Cu Viet Dung, Nguyen Huu Quynh, Ngo Quoc Tao, Tran Thi Minh Thu, “A images retrieval method base representation and manifold learning for dimensionality reduction with information from users”, Proceeding of FAIR 2019, ISBN: 978-604-913-867-6, Hà Nội – 2019.
Report overview Research improves effectiveness of image retrieval through distance learning.
1)    Research reports, proposals Research improves effectiveness of image retrieval through distance learning.
2)    Test program
Develop a test program of distance learning and feedback related to  the internationally commonly used Corel, including: 31,695 images Reported program test results.
3)    Report summarizes the implementation results of the topic
4)    Scientific articles (evidenced in the report on the implementation results of the topic)
5)    Support to train 02 PhD student and 04 graduate students (demonstrated  in the report on the implementation results of the topic)

It is possible to develop a application for retrieving landscape images, medicinal plants, fruits and beautiful place images,.