Download presentation
Presentation is loading. Please wait.
Published byCarson Moreton Modified over 9 years ago
1
MIT CSAIL Vision interfaces Towards efficient matching with random hashing methods… Kristen Grauman Gregory Shakhnarovich Trevor Darrell
2
MIT CSAIL Vision interfaces Motivation: Content-based image retrieval Data set of 30 scenes in Boston 1,079 database images 89 query images Features: Harris-Affine detector (max m=3,595) MSER detector (max m=1,707) SIFT-PCA descriptors Query
3
MIT CSAIL Vision interfaces Content-based image retrieval Pyramid match: ~1 second / query Optimal match: ~2 hours / query Number top retrievals Accuracy Even this is far too slow for any web-scale application!
4
MIT CSAIL Vision interfaces Sub-linear time image search N << N h 0111101 0110111 0110101 Randomized hashing techniques useful for sub-linear query time of very large image databases N Linear scan
5
MIT CSAIL Vision interfaces Pyramid match hashing For fixed-size sets, Locality-Sensitive Hashing [Indyk & Motwani 1998] provides bounded approximate similarity search over bijective matching [Indyk & Thaper 2003]; [Grauman & Darrell CVPR 2004, 2005] For varying set sizes, embedding of pyramid match (with product normalization) makes random hyperplane hashing possible under set intersection hash family of [Charikar 2002]. [Grauman PhD 2006]
6
MIT CSAIL Vision interfaces
7
MIT CSAIL Vision interfaces
8
MIT CSAIL Vision interfaces
9
MIT CSAIL Vision interfaces
10
MIT CSAIL Vision interfaces Single Frame Pose Estimation via Approximate Nearest Neighbor regression Obtain large DB of pose-appearance mappings Exploit fast methods for approximate nearest neighbor search in high dim. spaces. (e.g., LSH [Indyk and Motwani ‘98-’00]. )
11
MIT CSAIL Vision interfaces Approximate nearest neighbor techniques … … … Rendered (& hashed) Pose DB input Hash fcns. similar examples fall into same bucket in one or more hash table
12
MIT CSAIL Vision interfaces Single Frame Pose Estimation via Approximate Nearest Neighbor regression Render large DB of pose-appearance mappings Exploit fast methods for approximate nearest neighbor search in high dim. spaces. (e.g., LSH [Indyk and Motwani ‘98-’00]. ) Problem: signal distance dominated by nuisance variables Idea: find embedding (i.e., hash functions for LSH) most relevant to parameter (pose) similarity… [Shakhnarovich et. al ’03, Shakhnarovich ‘05]
13
MIT CSAIL Vision interfaces Pose estimation and Similarity-sensitive hashing … … … Rendered (& hashed) Pose DB input Pose- sensitive Hash fcns. NN similar in pose, not image [Shakhnarovich et. al ’03, Shakhnarovich ‘05]
14
MIT CSAIL Vision interfaces SSE / BoostPro Similarity Sensitive Embedding - Compute embedding H: I {0, 1} N such that | H(I( 1 )) - H(I( 2 )) | is small if 1 is close to 2 | H(I( 1 )) - H(I( 2 )) | is large otherwise - Use the embedding with approximate nearest neighbors retrieval (LSH) - Find H by training boosted classifier to learn “same-pair” and concatenate resulting weak learners … [Shakhnarovich 2005]
15
MIT CSAIL Vision interfaces PSH results ~200,000 examples in DB; 2 sec [Shakhnarovich et al. 2003, 2005]
16
MIT CSAIL Vision interfaces Conclusions Random Hashing techniques allow broad search; well suited for very high dimensional spaces Useful in domains where there is no prior knowledge about how to cluster or model data… Similarity (parameter) sensitive hashing can find distance related to task…effectively learn problem dependent distance measure and efficient means to index.
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.