Kapitel 14 Recognition Scene understanding / visual object categorization Pose clustering Object recognition by local features Image categorization Bag-of-features models Large-scale image search TexPoint fonts used in EMF. Read the TexPoint manual before you delete this box.: AA

Scene understanding (1) Scene categorization outdoor/indoor city/forest/factory/etc.

Scene understanding (2) Object detection find pedestrians

Scene understanding (3) sky mountain building tree building banner street lamp market people

Visual object categorization (1) ~10,000 to 30,000

Visual object categorization (2)

Visual object categorization (3)

Visual object categorization (4) Recognition is all about modeling variability camera position illumination shape variations within-class variations

Visual object categorization (5) Within-class variations (why are they chairs?)

Pose clustering (1) Working in transformation space - main ideas: Generate many hypotheses of transformation image vs. model, each built by a tuple of image and model features Correct transformation hypotheses appear many times Main steps: Quantize the space of possible transformations For each tuple of image and model features, solve for the optimal transformation that aligns the matched features Record a “vote” in the corresponding transformation space bin Find "peak" in transformation space

Pose clustering (2) Example: Rotation only. A pair of one scene segment and one model segment suffices to generate a transformation hypothesis.

Pose clustering (3) C.F. Olson: Efficient pose clustering using a randomized algorithm. IJCV, 23(2): 131-147, 1997. 2-tuples of image and model corner points are used to generate hypotheses. (a) The corners found in an image. (b) The four best hypotheses found with the edges drawn in. The nose of the plane and the head of the person do not appear because they were not in the models.

Object recognition by local features (1) D.G. Lowe: Distinctive image features from scale-invariant keypoints. IJCV, 60(2): 91-110, 2004 The SIFT features of training images are extracted and stored For a query image Extract SIFT features Efficient nearest neighbor indexing Pose clustering by Hough transform For clusters with >2 keypoints (object hypotheses): determine the optimal affine transformation parameters by least squares method; geometry verification Input Image Stored Image

Object recognition by local features (2) Cluster of 3 corresponding feature pairs

Object recognition by local features (3)

Object recognition by local features (4)

Image categorization (1) Training Labels Training Images Training Image Features Classifier Training Trained Classifier Testing Image Features Trained Classifier Prediction Outdoor Test Image

Image categorization (2) Global histogram (distribution): color, texture, motion, … histogram matching distance

Image categorization (3) Cars found by color histogram matching See Chapter “Inhaltsbasierte Suche in Bilddatenbanken

Bag-of-features models (1) Bag of ‘words’ Object

Bag-of-features models (2) Origin 1: Text recognition by orderless document representation (frequencies of words from a dictionary), Salton & McGill (1983) US Presidential Speeches Tag Cloud http://chir.ag/phernalia/preztags/

Bag-of-features models (3) Origin 2: Texture recognition Texture is characterized by the repetition of basic elements or textons For stochastic textures, it is the identity of the textons, not their spatial arrangement, that matters

Bag-of-features models (4) histogram Universal texton dictionary

Bag-of-features models (5) We need to build a “visual” dictionary! G. Cruska et al.: Visual categorization with bags of keypoints. Proc. of ECCV, 2004.

Bag-of-features models (6) Main step 1: Extract features (e.g. SIFT) …

Bag-of-features models (7) Main step 2: Learn visual vocabulary (e.g. using clustering) clustering

Bag-of-features models (8) visual vocabulary (cluster centers)

Bag-of-features models (9) Example: learned codebook … Appearance codebook Source: B. Leibe

Bag-of-features models (10) Main step 3: Quantize features using “visual vocabulary”

Bag-of-features models (11) Main step 4: Represent images by frequencies of “visual words” (i.e., bags of features)

Bag-of-features models (12) Example: representation based on learned codebook

Bag-of-features models (13) Main step 5: Apply any classifier to the histogram feature vectors

Bag-of-features models (14) Example:

Bag-of-features models (15) Caltech6 dataset Dictionary quality and size are very important parameters!

Bag-of-features models (16) Caltech6 dataset Action recognition J.C. Niebles, H. Wang, L. Fei-Fei: Unsupervised learning of human action categories using spatial-tempotal words. IJCV, 79(3): 299-318, 2008.

Large-scale image search (1) J. Philbin, et al.: Object retrieval with large vocabularies and fast spatial matching. Proc. of CVPR, 2007 Query Results from 5k Flickr images

Large-scale image search (2) Aachen Cathedral [Quack, Leibe, Van Gool, CIVR’08] Mobile tourist guide self-localization object/building recognition photo/video augmentation

Large-scale image search (3)

Large-scale image search (4) Moulin Rouge Old Town Square (Prague) Tour Montparnasse Colosseum Viktualienmarkt Maypole Application: Image auto-annotation Left: Wikipedia image Right: closest match from Flickr

Sources K. Grauman, B. Leibe: Visual Object Recognition. Morgen & Claypool Publishers, 2011 R. Szeliski: Computer Vision: Algorithms and Applications. Springer, 2010. (Chapter 14 “Recognition”) Course materials from others (G. Bebis, J. Hays, S. Lazebnik, …)