the patches by d-dimensional vectors ③Make correspondences between similar patches ④Calculate similarity between the images Similarity: 3 Position (x, y) Orientation θ Scale σ Feature vector f (e.g., 128-dim SIFT) Local feature
than Euclidean distance in comparing histograms such as SIFT • Hellinger kernel (Bhattacharyyaʼs coefficient) for L1 normalized histograms x and y: • Explicit feature map of x into xʼ : – L1 normalize x – element-wise square root x to give xʼ – then xʼ is L2 normalized • Computing Euclidean distance in the feature map space is equivalent to Hellinger distance in the original space: RootSIFT RootSIFT
a large number of training vectors – Perform clustering algorithm (e.g., k-means) – Centroids of clusters = visual words (VWs) • Online: – All features are assigned to their nearest visual words – An image is represented by the frequency histogram of VWs – (Dis)similarity is defined by the distance between histograms Visual words (VW) VW1 VWn VW2 … Visual words - - " " " - - - " " " - - - " " " - - - " " " - - - " " " - Frequency } 1 | { N i i £ £ = v V
VWn ・ ・ ・ ・ ・ ・ Indexing step (quantization) Search step (quantization) Match Match Matching can be performed in O(1) with an inverted index Query image Reference images Nearest VW
(m < 50) verified results of original query • Construct new query using average of these results Without geometric verification, QE degrades accuracy! Query image Verified results New query
image ROI ROI ROI ROI ROI ROI First verified results ROI ROI ROI ROI ROI ROI • Calculate relative change in resolution • Construct average query for each resolution New query1 New query2 New query3
linear SVM classifier – Use verified results as positive training data – Use low ranked images as negative training data – Rank images on their signed distance from the decision boundary – Reranking can be efficient with an inverted index!