Learning View-Model Joint Relevance for 3D Object Retrieval
read more
Citations
Multi-Modal Clique-Graph Matching for View-Based 3D Model Retrieval
Local Bit-Plane Decoded Pattern: A Novel Feature Descriptor for Biomedical Image Retrieval
Exploring Deep Learning for View-Based 3D Model Retrieval
Multi-view ensemble manifold regularization for 3D object recognition
View-Based 3-D Model Retrieval: A Benchmark
References
A Comparison of Document Clustering Techniques
Using spin images for efficient object recognition in cluttered 3D scenes
Topology matching for fully automatic similarity estimation of 3D shapes
Invariant image recognition by Zernike moments
Shape distributions
Related Papers (5)
On Visual Similarity Based 3D Model Retrieval
Enhanced Local Texture Feature Sets for Face Recognition Under Difficult Lighting Conditions
Frequently Asked Questions (15)
Q2. What are the common 3D shape descriptors?
In transform-based method, transform coefficients are employed as the 3D shape descriptor, such as 3D Fourier [26], spherical trace transform [27], radialized extend function [28], and concrete radialized shperical projection [29].
Q3. What is the common method used to measure the relevance between a target object and the query?
In [41], a positive matching model and a negative matching model were used to measure the relevance between a target object and the query.
Q4. What are the methods used to represent 3D objects?
To represent 3D objects, low-levelfeatures, such as volumetric descriptor [11] and surface geometry [8], [12], and high-level features, such as the method in [14] were employed in previous works.
Q5. What is the regularizer term on the view-based hypergraphstructure?
(4) KV ( f ) is the regularizer term on the view-based hypergraphstructure, KM ( f ) is the regularizer term on the model-basedThe vertex degree matrix and the edge degree matrix canbe denoted by two diagonal matrices Dv and De.
Q6. What is the common method used to measure the distance between two 3D models?
In [42], curvature scale space was employed as the view descriptor, which was further combined with Zernike Moments to measure the distance between two 3D models.
Q7. What are the different types of 3D shape descriptors?
According to [22], 3D shape descriptors can be divided into four categories, i.e., histogram-based method [9], [23]–[25], transformbased method [26]–[29], graph-based method [30]–[32] and view-based method [21], [33], [34].
Q8. What datasets are used to evaluate the performance of the proposed method?
In their experiments, to evaluate the performance of the proposed method, three datasets are employed, i.e., National Taiwan University 3D Model database (NTU) [33], Princeton Shape Benchmark (PSB) [19] and Shape Retrieval Content 2009 (SHREC) [2]
Q9. How can the object relevance vector be obtained?
With the learned combination weights, the model-based and view-based data can be optimally explored simultaneously and the relevance vector f can be obtained.
Q10. What is the method used to generate multiple views?
In the camera constraint free method (CCFV) [41], a set of representative views are selected from the originally captured multiple views via view clustering and a probabilistic matching method is then employed to calculate the similarity between each two 3D objects.
Q11. What is the common method of comparing 3D objects?
The comparison between 3D objects is formulated as a probabilistic approach to measure the posterior probability of the target object given the query.
Q12. What is the relevance of the 3D objects in HL?
In HL, the relevanceamong 3D objects is formulated in a hypergraph structure, where the hyperedges are generated using the view clustering results.
Q13. What is the objective function for the view-based hypergraph?
This objective function aims to minimize the empirical loss and the regularizers on the model-based graph and the view-based hypergraph simultaneously which can lead to the optimal relevance vector f for retrieval.
Q14. What is the meaning of the term "High level semantic space"?
With user feedback, these low level features are mapped to high level semantic space, which is another Euclidean space and can be regarded as a dimension reduction or feature selection method.
Q15. What is the proposed method for 3D object retrieval?
Given the view information of 3D objects, the proposed method first constructs a hypergraph to formulate the relationship among 3D objects with the view connections.