Apictorial Jigsaw Puzzles: The Computer Solution of a Problem in Pattern Recognition
TL;DR: This paper describes the development of a procedure that enables a digital computer to solve ``apictorial'' jigsaw puzzles, i.e., puzzles in which all pieces are uniformly gray and the only available information is the shape of the pieces.
Abstract: This paper describes the development of a procedure that enables a digital computer to solve ``apictorial'' jigsaw puzzles, i.e., puzzles in which all pieces are uniformly gray and the only available information is the shape of the pieces. The problem was selected because it provided an excellent vehicle to develop computer techniques for manipulation of arbitrary geometric patterns, for pattern identification, and for game solving. The kinds of puzzles and their properties are discussed in detail. Methods are described for characterizing and classifying piece contours, for selecting and ordering pieces that are ``most likely'' to mate with a given piece, for determining likelihood of fit, for overcoming ambiguities, and for evaluation of the progressive puzzle assembly. An illustration of an actual computer solution of a puzzle is given.
Citations
More filters
08 Oct 2016
TL;DR: In this article, a siamese-ennead convolutional neural network (CFN) is proposed to build features suitable for object detection and classification without human annotation and later transferred via fine-tuning on a different, smaller and labeled dataset.
Abstract: We propose a novel unsupervised learning approach to build features suitable for object detection and classification. The features are pre-trained on a large dataset without human annotation and later transferred via fine-tuning on a different, smaller and labeled dataset. The pre-training consists of solving jigsaw puzzles of natural images. To facilitate the transfer of features to other tasks, we introduce the context-free network (CFN), a siamese-ennead convolutional neural network. The features correspond to the columns of the CFN and they process image tiles independently (i.e., free of context). The later layers of the CFN then use the features to identify their geometric arrangement. Our experimental evaluations show that the learned features capture semantically relevant content. We pre-train the CFN on the training set of the ILSVRC2012 dataset and transfer the features on the combined training and validation set of Pascal VOC 2007 for object detection (via fast RCNN) and classification. These features outperform all current unsupervised features with \(51.8\,\%\) for detection and \(68.6\,\%\) for classification, and reduce the gap with supervised learning (\(56.5\,\%\) and \(78.2\,\%\) respectively).
1,808 citations
Posted Content•
TL;DR: A novel unsupervised learning approach to build features suitable for object detection and classification and to facilitate the transfer of features to other tasks, the context-free network (CFN), a siamese-ennead convolutional neural network is introduced.
Abstract: In this paper we study the problem of image representation learning without human annotation. By following the principles of self-supervision, we build a convolutional neural network (CNN) that can be trained to solve Jigsaw puzzles as a pretext task, which requires no manual labeling, and then later repurposed to solve object classification and detection. To maintain the compatibility across tasks we introduce the context-free network (CFN), a siamese-ennead CNN. The CFN takes image tiles as input and explicitly limits the receptive field (or context) of its early processing units to one tile at a time. We show that the CFN includes fewer parameters than AlexNet while preserving the same semantic learning capabilities. By training the CFN to solve Jigsaw puzzles, we learn both a feature mapping of object parts as well as their correct spatial arrangement. Our experimental evaluations show that the learned features capture semantically relevant content. Our proposed method for learning visual representations outperforms state of the art methods in several transfer learning benchmarks.
1,571 citations
Cites background from "Apictorial Jigsaw Puzzles: The Comp..."
...There is also a sizeable literature on solving Jigsaw puzzles computationally (see, for example, [27,12,26])....
[...]
TL;DR: Various forms of line drawing representation are described, different schemes of quantization are compared, and the manner in which a line drawing can be extracted from a tracing or a photographic image is reviewed.
Abstract: This paper describes various forms of line drawing representation, compares different schemes of quantization, and reviews the manner in which a line drawing can be extracted from a tracing or a photographic image. The subjective aspects of a line drawing are examined. Different encoding schemes are compared, with emphasis on the so-called chain code which is convenient for highly irregular line drawings. The properties of chain-coded line drawings are derived, and algorithms are developed for analyzing line drawings to determine various geometric features. Procedures are described for rotating, expanding, and smoothing line structures, and for establishing the degree of similarity between two contours by a correlation technique. Three applications are described in detail: automatic assembly of jigsaw puzzles, map matching, and optimum two-dimensional template layout
1,485 citations
01 Jan 1969
TL;DR: The field of picture processing by computer is reviewed from a technique-oriented standpoint and the processing of given pictures (as opposed to computer-synthesized pictures) is considered.
Abstract: : The field of picture processing by computer is reviewed from a technique-oriented standpoint. Only the processing of given pictures (as opposed to computer-synthesized pictures) is considered. Specific areas covered include: (a) Pictures as information sources and their efficient encoding; (b) Approximation of pictures - sampling and quantization techniques; (c) Position-invariant operations on pictures and their implementation (digital, electro-optical, optical); applications to matched filtering (template matching), spatial frequency filtering and image restoration, measurement of image quality, and image enhancement ('smoothing' and 'sharpening'); (d) Picture properties (linear; local and 'textural'; random) useful for pictorial pattern recognition; (e) 'Figure extraction' from pictures; figure properties (topology, size, shape); (f) Picture description and 'picture languages.' (Author)
712 citations
15 Jun 2019
TL;DR: This model learns the semantic labels in a supervised fashion, and broadens its understanding of the data by learning from self-supervised signals how to solve a jigsaw puzzle on the same images, which helps the network to learn the concepts of spatial correlation while acting as a regularizer for the classification task.
Abstract: Human adaptability relies crucially on the ability to learn and merge knowledge both from supervised and unsupervised learning: the parents point out few important concepts, but then the children fill in the gaps on their own. This is particularly effective, because supervised learning can never be exhaustive and thus learning autonomously allows to discover invariances and regularities that help to generalize. In this paper we propose to apply a similar approach to the task of object recognition across domains: our model learns the semantic labels in a supervised fashion, and broadens its understanding of the data by learning from self-supervised signals how to solve a jigsaw puzzle on the same images. This secondary task helps the network to learn the concepts of spatial correlation while acting as a regularizer for the classification task. Multiple experiments on the PACS, VLCS, Office-Home and digits datasets confirm our intuition and show that this simple method outperforms previous domain generalization and adaptation solutions. An ablation study further illustrates the inner workings of our approach.
678 citations
Cites background from "Apictorial Jigsaw Puzzles: The Comp..."
...In the area of computer science and artificial intelligence it was first introduced by [17], which proposed a 9-piece puzzle solver based only on shape information and ignoring the image content....
[...]
References
More filters
TL;DR: In this article, two machine learning procedures have been investigated in some detail using the game of checkers, and enough work has been done to verify the fact that a computer can be programmed so that it will lear...
Abstract: Two machine-learning procedures have been investigated in some detail using the game of checkers. Enough work has been done to verify the fact that a computer can be programmed so that it will lear...
2,845 citations
TL;DR: It is shown that one can determine through the use of relatively simple numerical techniques whether a given arbitrary plane curve is open or closed, whether it is singly or multiply connected, and what area it encloses.
Abstract: A method is described which permits the encoding of arbitrary geometric configurations so as to facilitate their analysis and manipulation by means of a digital computer. It is shown that one can determine through the use of relatively simple numerical techniques whether a given arbitrary plane curve is open or closed, whether it is singly or multiply connected, and what area it encloses. Further, one can cause a given figure to be expanded, contracted, elongated, or rotated by an arbitrary amount. It is shown that there are a number of ways of encoding arbitrary geometric curves to facilitate such manipulations, each having its own particular advantages and disadvantages. One method, the so-called rectangular-array type of encoding, is discussed in detail. In this method the slope function is quantized into a set of eight standard slopes. This particular representation is one of the simplest and one that is most readily utilized with present-day computing and display equipment.
1,751 citations
01 Jan 1961
TL;DR: The problems of heuristic programming can be divided into five main areas: Search, Pattern-Recognition, Learning, Planning, and Induction as discussed by the authors, and the most successful heuristic (problem-solving) programs constructed to date.
Abstract: The problems of heuristic programming-of making computers solve really difficult problems-are divided into five main areas: Search, Pattern-Recognition, Learning, Planning, and Induction. A computer can do, in a sense, only what it is told to do. But even when we do not know how to solve a certain problem, we may program a machine (computer) to Search through some large space of solution attempts. Unfortunately, this usually leads to an enormously inefficient process. With Pattern-Recognition techniques, efficiency can often be improved, by restricting the application of the machine's methods to appropriate problems. Pattern-Recognition, together with Learning, can be used to exploit generalizations based on accumulated experience, further reducing search. By analyzing the situation, using Planning methods, we may obtain a fundamental improvement by replacing the given search with a much smaller, more appropriate exploration. To manage broad classes of problems, machines will need to construct models of their environments, using some scheme for Induction. Wherever appropriate, the discussion is supported by extensive citation of the literature and by descriptions of a few of the most successful heuristic (problem-solving) programs constructed to date.
1,318 citations