Efficient Visual Search for Objects in Videos JOSEF SIVIC AND ANDREW ZISSERMAN PRESENTERS: ILGE AKKAYA & JEANNETTE CHANG MARCH 1, 2011
Download ReportTranscript Efficient Visual Search for Objects in Videos JOSEF SIVIC AND ANDREW ZISSERMAN PRESENTERS: ILGE AKKAYA & JEANNETTE CHANG MARCH 1, 2011
Efficient Visual Search for Objects in Videos
J O S E F S I V I C A N D A N D R E W Z I S S E R M A N P R E S E N T E R S : I L G E A K K A Y A & J E A N N E T T E C H A N G M A R C H 1 , 2 0 1 1
Introduction
Text Query Image Query
Generalize text retrieval methods to
Results: Documents Results: Frames
non-textual information
State-of-the-Art before this paper…
Text-based search for images (Google Images) Object recognition Barnard, et al. (2003): “Matching words and pictures” Sivic, et al. (2005): “Discovering objects and their location in images” Sudderth, et al. (2005): “Learning hierarchical models of scenes, objects, and parts” Scene classification Fei-Fei and Perona (2005): “A Bayesian hierarchical model for learning natural scene categories” Quelhas, et al. (2005): “Modeling scenes with local descriptors and latent aspects” Lazebnik, et al. (2006): “Beyond bag of features: Spatial pyramid matching for recognizing natural scene categories”
Introduction (cont.)
Retrieve specific objects vs. categories of objects/scenes (“Camry” logo vs. cars) Employ text retrieval techniques for visual search, with images as queries and results Why Text Retrieval Approach?
Matches essentially precomputed so that no delay at run time Any object in video can be retrieved without modification of descriptors originally built for video
Overview of the Talk
Visual Search Algorithm
Offline Pre-Processing
Real-Time Query A Few Implementation Details
Performance
General Results Testing Individual Words Using External Images As Queries
A Few Challenges and Future Directions Concluding Remarks Demo of the Algorithm
Overview of the Talk
Visual Search Algorithm
Offline Pre-Processing
Real-Time Query A Few Implementation Details
Performance
General Results Testing Individual Words Using External Images As Queries
A Few Challenges and Future Directions Concluding Remarks Demo of the Algorithm
Pre-Processing (Offline)
Detection of Affine Covariant Regions
Typically ~1200 regions / frame (720x576) Elliptical regions Each region represented by 128-dimensional SIFT vector SIFT features provide invariance against affine transformations
Two types of affine covariant regions: 1. Shape-Adapted(SA):
Mikolajczyk et al.
Elliptical Shape adaptation about a Harris interest point Often centered on corner-like features 1. Maximally-Stable(MS):
Proposed by Matas et al.
Intensity watershed image segmentation High-contrast blobs
Pre-Processing (Offline)
Tracking regions through video and rejecting unstable regions
Any region that does not survive for 3+ frames is rejected These regions are not potentially interesting Reduces number of regions/frame to approx. 50% (~600/frame)
Pre-Processing (Offline)
Visual Indexing Using Text-Retrieval Methods
TEXT
Represent words by the “stems” ‘write’ ‘writing’ ‘write’ ‘written’
mapped to
Stop-list common words ‘a/an/the’ Rank search results according to how close the query words occur within retrieved document
IMAGE
Cluster similar regions into ‘visual words’ Stop-list common visual words Use spatial information to check retrieval consistency
Visual Vocabulary
Purpose: Cluster regions from multiple frames into fewer groups called ‘visual words’ Each descriptor: 128-vector K-means clustering (explain more) ~300K descriptors mapped into 16K visual words (600 regions/frame x ~500 frames) (6000 SA, 10000 MS regions used)
K-Means Clustering
Purpose: Cluster N data points (SIFT descriptors) into K clusters (visual words) K = desired number of cluster centers (mean points) Step 1: Randomly guess K mean points
Step 2: Calculate nearest mean point to assign each data point to a cluster center In this paper, Mahalanobis distance is used to determine ‘nearest cluster center’.
d
(
x
1 ,
x
2 ) = (
x
1 -
x
2 )
T
S 1 (
x
1 -
x
where ∑ is the covariance matrix for all descriptors, x 2 is the length 128 mean vector and x 1 ’s are the descriptor vectors(i.e. data points) 2 )
Step 3: Recalculate cluster centers and distances, repeat until stationarity
Examples of Clusters of Regions
Samples of normalized affine covariant regions
Pre-Processing (Offline)
Remove Stop-Listed Words
Analogy to text-retrieval: ‘a’, ‘and’, ‘the’ … are not distinctive words Common words cause mismatches 5-10% of the most common visual words are stopped 800-1600 / 16000 words are stopped (Upper row) Matches before stop listing (Lower row) Matches after stop listing
Pre-Processing (Offline)
tf-idf Weighting
(term frequency-inverse document frequency weighting) n id n d : #of occurrences of word(visual word) i in document(frame) d : total number of words in document d N i : total number of documents containing term I N : number of documents in the database t i : weighted word frequency
Each document(frame) represented by: where v = number of visual words in vocabulary And v d = the tf-idf vector of the particular frame d
Inverted File Indexing
Visual Word Index
1 2 … N
Found in Frames:
1,4,5 1,2,10 …
Overview of the Talk
Visual Search Algorithm
Offline Pre-Processing
Real-Time Query A Few Implementation Details
Performance
General Results Testing Individual Words Using External Images As Queries
A Few Challenges and Future Directions Concluding Remarks Demo of the Algorithm
Real-Time Query
1.
2.
3.
Determine the set of visual words found within the query region Retrieve keyframes based on visual word frequencies (Ns = 500) Re-rank retrieved keyframes using spatial consistency
Retrieve keyframes based on visual word frequencies v q : vector containing visual word frequencies corresponding to query region is computed the normalized scalar product of v q computed: with v d ’s are
Spatial Consistency Voting
Analogy: Google text document retrieval Matched covariant regions in the retrieved frames should have a similar spatial arrangement Search area: 15 nearest spatial neighbors of each match Each neighboring region which also matches in the retrieved frame, votes for the frame
Spatial Consistency Voting
Matched pair of words (A,B) Each region in defined search area in both frames casts a vote For the match (A,B) (upper row)Matches after stop-listing (lower row) Remaining matches after spatial consistency voting
Query Frame Sample Retrieved Frame 1: 2: 3-4: 5-6: 7-8: Query Region Close-up version of 1 Initial matches Matches after stop-listing Matches after spatial consistency matching 1 3 5 7 2 4 6 8
Overview of the Talk
Visual Search Algorithm
Offline Pre-Processing
Real-Time Query A Few Implementation Details
Performance
General Results Testing Individual Words Using External Images As Queries
A Few Challenges and Future Directions Concluding Remarks Demo of the Algorithm
Implementation Details
Offline Processing: 100-150K frames/typical feature length film, Refined to 4000-6000 keyframes Descriptors are computed for stable regions in each frame Each region is assigned to a visual word Visual words over all keyframes assembled into an inverted file-structure
Algorithm Implementation
Real-Time Process: User selects query region Visual words are identified within query region A short list of Ns = 500 keyframes retrieved based on tf-idf vector similarity Similarity is recomputed considering spatial consistency voting
Example Visual Search
Overview of the Talk
Visual Search Algorithm
Offline Pre-Processing
Real-Time Query A Few Implementation Details
Performance
General Results Testing Individual Words Using External Images As Queries
A Few Challenges and Future Directions Concluding Remarks Demo of the Algorithm
Query Image A Few Retrieved Matches
Retrieval Examples
Query Image
Retrieval Examples (cont.)
A Few Retrieved Matches
Performance of the Algorithm
Tried 6 object queries (1) Red Clock (2) Black Clock (3) “Frame’s” Sign (4) Digital Clock (5) “Phil” Sign (6) Microphone
Performance of the Algorithm (cont.)
Evaluated on the level of shots rather than keyframes Measured using precision-recall plots 𝑃𝑟𝑒𝑐𝑖𝑠𝑖𝑜𝑛 = # 𝑜𝑓 𝑐𝑜𝑟𝑟𝑒𝑐𝑡𝑙𝑦 𝑟𝑒𝑡𝑟𝑖𝑒𝑣𝑒𝑑 𝑠ℎ𝑜𝑡𝑠 𝑇𝑜𝑡𝑎𝑙 # 𝑜𝑓 𝑠ℎ𝑜𝑡𝑠 𝒓𝒆𝒕𝒓𝒊𝒆𝒗𝒆𝒅 # 𝑜𝑓 𝑐𝑜𝑟𝑟𝑒𝑐𝑡𝑙𝑦 𝑟𝑒𝑡𝑟𝑖𝑒𝑣𝑒𝑑 𝑠ℎ𝑜𝑡𝑠 𝑅𝑒𝑐𝑎𝑙𝑙 = 𝑇𝑜𝑡𝑎𝑙 # 𝑜𝑓 𝑠ℎ𝑜𝑡𝑠 𝒘𝒊𝒕𝒉 𝒐𝒃𝒋𝒆𝒄𝒕 Precision like measure of fidelity or exactness Recall like measure of completeness
Performance of the Algorithm (cont.)
Ideally, precision = 1 for all recall values Average Precision (AP) , ideally AP = 1
Examples of Missed Shots
Extreme viewing angles Original query object Low-ranked shot
Examples of Missed Shots (cont.)
Significant changes in scale and motion blurring Original query object Low-ranked shot
Qualitative Assessment of Performance
General trends Higher precision at low recall levels Bias towards lightly textured regions detectable by SA/MS detectors Could address these challenges by adding more covariant regions Other Difficulties Textureless regions (e.g., mug) Thin or wiry objects (e.g., bike) Highly-deformable (e.g., clothing)
Overview of the Talk
Visual Search Algorithm
Offline Pre-Processing
Real-Time Query A Few Implementation Details
Performance
General Results Testing Individual Words Using External Images As Queries
A Few Challenges and Future Directions Concluding Remarks Demo of the Algorithm
Quality of Individual Visual Words
Using single visual word as query Tests the expressiveness of the visual vocabulary Sample query Given an object of interest, select one of the visual words from that object Retrieve all frames that contain the visual word (no ranking) Retrieval considered correct if contains object of interest
Examples of Individual Visual Words
Top row: Scale-normalized close-ups of elliptical regions overlaid on query image Bottom row: Corresponding normalized regions
Results of Individual Word Searches
Individual words are “noisy” Intuitively because words occur in multiple objects and do not cover all occurrences of the object
Quality of Individual Visual Words
Unrealistic
Require each word to occur on only one object (high precision) Growing number of objects would result in growing number of words
Realistic
Visual words shared across objects, with objects represented by a combination of words
Overview of the Talk
Visual Search Algorithm
Offline Pre-Processing
Real-Time Query A Few Implementation Details
Performance
General Results Testing Individual Words Using External Images As Queries
A Few Challenges and Future Directions Concluding Remarks Demo of the Algorithm
Searching for Objects From Outside of the Movie
Used external query images from the internet Manually labeled all occurrences of external queries in movies Results
External Query Image
Sony logo Hollywood sign Notre Dame
No. of Occurrences
3 1 1
Rankings of Retrieved Occurrences
1 st , 4 th , 35 th 1 st 1 st
AP (Average Precision)
0.53
1 1
Sample External Query Results
Potential Applications
Overview of the Talk
Visual Search Algorithm
Offline Pre-Processing
Real-Time Query A Few Implementation Details
Performance
General Results Testing Individual Words Using External Images As Queries
A Few Challenges and Future Directions Concluding Remarks Demo of the Algorithm
Challenge I: Visual Vocabularies for Very Large Scale Retrieval
Current progress: 150000 frame feature movie reduced to 6000 keyframes and then processed Ultimate goal: indexing billions of online images to build a visual search engine
(a) (c) external images downloaded from the Internet (b) Correct retrieval frame from the movie ‘Pretty Woman’ (d) Correct retrieval from the movie ‘Charade’ Should the vocabulary increase in size as the image archive grows?
How discriminative should the words be?
Generalization of images from one movie to an outside database of images?
Learning a universal visual vocabulary still remains a challenge
Challenge II: Retrieval of 3D Objects
Current algorithm covers successful detection despite slight changes in viewpoint, illumination, partial occlusion due to SIFT features However, 3D retrieval is fundamentally a bigger challenge
Proposed approach 1:
Automatic association of images using temporal information Grouping front-side-back of a car in a video Possible either in query and/or database side Query-Side Matching: Associated query frames are computed and used for 3D image search Query-Side matching of associated frames
Proposed approach 1 (cont.)
Grouping on database side: Query on a single aspect is expected to retrieve pregrouped frames associated with 3D image (Top Row) Query image (Bottom Rows) Matching frames
Proposed approach 2:
Building an explicit 3-D model for each 3-D object in the Video Focus is more on model building than detection Only rigid objects considered
Challenge III: Verification using Spatial Structure
Spatial consistency was helpful, but could be improved A few suggestions Caution with using measures for rigid geometry Reduce cost using hierarchical approach Two complementary methods Ferrari et al. (2004): matching deformable objects Rothganger et al. (2003): matching 3D objects
Verification Using Spatial Structure (cont.)
Method 1 (Ferrari) Based on spatial overlap of local regions Requires regions to match individually and pattern of intersection between neighboring regions to be preserved Performance Pro: Works well with deformations Con: Computationally expensive
Verification Using Spatial Structure (cont.)
Method 2 (Rothganger) Based on 3-D object model Requires consistency of local appearance descriptors and geometric consistency Performance Pro: Object can be matched in diverse (even novel) poses Con: 3-D model built offline, requires up to 20 images of object taken from different viewpoints
Overview of the Talk
Visual Search Algorithm
Offline Pre-Processing
Real-Time Query A Few Implementation Details
Performance
General Results Testing Individual Words Using External Images As Queries
A Few Challenges and Future Directions Concluding Remarks Demo of the Algorithm
Conclusion
Demonstrated scalable object retrieval architecture which uses Visual vocabulary based on vector-quantized viewpoint invariant descriptors Efficient indexing techniques from text retrieval A few notable differences between document and image bag-of-words retrieval Spatial information Numbers of “words” in query Matching requirements
Looking forward…
TinEye (May 2008) Image-based search engine Given a query image, searches for altered versions of that image (scaled or cropped) 1.86 billion images indexed Google Goggles (2009) Use phone to take photo, results from the internet Limited categories
Overview of the Talk
Visual Search Algorithm
Offline Pre-Processing
Real-Time Query A Few Implementation Details
Performance
General Results Testing Individual Words Using External Images As Queries
A Few Challenges and Future Directions Concluding Remarks Demo of the Algorithm
Demo of Retrieval Algorithm
Live demonstration
Main References
D. Lowe. Distinctive Image Features from Scale- Invariant Keypoints. International Journal of Computer Vision. 2(60):91.110, 2004.
J. Sivic and A. Zisserman. Efficient visual search for objects in videos. Proc. IEEE, 96(4):548–566, 2008.
W. Qian “Video Google: A Text Retrieval Approach to Object Matching in Videos.” www.mriedel.ece.umn.edu/wiki/index.php/Weikang_Qian