Agenda Introduction Bag-of-words models Visual words with spatial location Part-based models Discriminative methods Segmentation and recognition Recognition-based image retrieval Datasets & Conclusions
Retrieval domains Internet image search Video search for people/objects Searching home photo collections
Learning from Internet Image Search Joint learning of text and images Large scale retrieval
Noisy labels
Improving Google’s Image Search Fergus, Fei-Fei, Perona, Zisserman, ICCV 2005 Variant of pLSA that includes spatial information
Topics in model Re-ranking result: Motorbike Automatically chosen topic
Animals on the Web Berg and Forsyth, CVPR 2006 Gather images using text search Use LDA to discover “good” images using features based on nearby text, shape, color
Boostrapping of Image Search 2 4 Images returned with PENGUIN query Removal of drawings and abstract images Naives Bayes ranking using noisy metadata Train SVM……. Schroff, Zisserman, Criminisi, Harvesting Image Databases from the Web, ICCV 2007 Final ranking using SVM
OPTIMOL Li, Wang, Fei-Fei CVPR 07
Learning from Internet Image Search Joint learning of text and images Large scale retrieval
Matching Words and Pictures Barnard, Duygulu, de Freitas, Forsyth, Blei, Jordan, JMLR 2003
Text to Images
Images to text Use Blobworld or nCuts to segments images into regions Need to deduce labels attached to each image
Images to text result
Names and Faces in the News Berg, Berg, Edwards, Maire, White, Teh, Learned-Miller, Forsyth. CVPR Find faces (standard face detector), rectify them to same pose. 2.Perform Kernel PCA and Linear Discriminant Analysis (LDA). 3.Extract names from text. 4.Cluster faces, with each name corresponding to a cluster. 5.Use language model to refine results Collected 500,000 images and text captions from Yahoo! News
Initial clusters
Clusters refined with language model
Learning from Internet Image Search Joint learning of text and images Large scale retrieval
Vocabulary tree Nistér & Stewénius CVPR KD-tree in descriptor space Inverse lookup of features Specific object recognition Not category-level
Slide from D. Nister
Pyramid Match Hashing Grauman & Darell, CVPR 2007 Combines Pyramid Match Kernel (efficient computation of correspondences between two set of vectors) with Locality Sensitive Hashing (LSH) [Indyk & Motwani 98] Allows matching of the set of features in a query image to sets of features in other images in time that is sublinear in # images Theoretical guarantees
Salakhutdinov and Hinton, SIGIR 2007 Torralba, Fergus, Weiss, CVPR 2008 Map images to compact binary codes Hash codes for fast lookup Semantic Hashing