Probability Distributions.- Linear Models for Regression.- Linear Models for Classification.- Neural Networks.- Kernel Methods.- Sparse Kernel Machines.- Graphical Models.- Mixture Models and EM.- Approximate Inference.- Sampling Methods.- Continuous Latent Variables.- Sequential Data.- Combining Models.

Pattern Recognition and Machine Learning

“Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks”の学習報告

Detection of visually salient image regions is useful for applications like object segmentation, adaptive compression, and object recognition. In this paper, we introduce a method for salient region detection that outputs full resolution saliency maps with well-defined boundaries of salient objects. These boundaries are preserved by retaining substantially more frequency content from the original image than other existing techniques. Our method exploits features of color and luminance, is simple to implement, and is computationally efficient. We compare our algorithm to five state-of-the-art salient region detection methods with a frequency domain analysis, ground truth, and a salient object segmentation application. Our method outperforms the five algorithms both on the ground-truth evaluation and on the segmentation task by achieving both higher precision and better recall.

/pdf/frequency-tuned-salient-region-detection-4po18ylvds.pdf

Frequency-tuned salient region detection

Automatic estimation of salient object regions across images, without any prior assumption or knowledge of the contents of the corresponding scenes, enhances many computer vision and computer graphics applications. We introduce a regional contrast based salient object detection algorithm, which simultaneously evaluates global contrast differences and spatial weighted coherence scores. The proposed algorithm is simple, efficient, naturally multi-scale, and produces full-resolution, high-quality saliency maps. These saliency maps are further used to initialize a novel iterative version of GrabCut, namely SaliencyCut, for high quality unsupervised salient object segmentation. We extensively evaluated our algorithm using traditional salient object detection datasets, as well as a more challenging Internet image dataset. Our experimental results demonstrate that our algorithm consistently outperforms 15 existing salient object detection and segmentation methods, yielding higher precision and better recall rates. We also show that our algorithm can be used to efficiently extract salient object masks from Internet images, enabling effective sketch-based image retrieval (SBIR) via simple shape comparisons. Despite such noisy internet images, where the saliency regions are ambiguous, our saliency guided image retrieval achieves a superior retrieval rate compared with state-of-the-art SBIR methods, and additionally provides important target object region information.

https://mmcheng.net/mftp/SalObj/PosterSaliency.pdf

Global contrast based salient region detection

From the Publisher: 
The accessible presentation of this book gives both a general view of the entire computer vision enterprise and also offers sufficient detail to be able to build useful applications. Users learn techniques that have proven to be useful by first-hand experience and a wide range of mathematical methods. A CD-ROM with every copy of the text contains source code for programming practice, color images, and illustrative movies. Comprehensive and up-to-date, this book includes essential topics that either reflect practical significance or are of theoretical importance. Topics are discussed in substantial and increasing depth. Application surveys describe numerous important application areas such as image based rendering and digital libraries. Many important algorithms broken down and illustrated in pseudo code. Appropriate for use by engineers as a comprehensive reference to the computer vision enterprise.

Computer vision : a modern approach = 计算机视觉 : 一种现代的方法

A number of cortex-like hierarchical models of object recognition have been proposed these years. In this paper, we improve them by introducing supervision during forming combined local features. The traditional cortex-like hierarchical models always contain three layers which imitate the functions of neurons in ventral visual stream of primates. The bottom layer detects orientation information in a local area. Then the middle layer combines these information to form combined features. Finally, the top layer integrates combined features to form global features which are input into a classifier. In these models, three stages to form global features are all unsupervised. The supervision procedure only occurs after global features are generated, which is implemented by the classifier. But we think the supervision should occurs earlier. For a particular object recognition task, the second stage of generating global features is also supervised because only task relevant combinations are useful. In our paper, we analyze why introducing supervision in this stage is necessary. And we explain task relevant combined local features can be extracted by some feature selection algorithms. We also apply this improved system to a series of object classification problems and compare it with traditional models. The simulation results show that our improvement really boosts object recognition performance.

/pdf/object-recognition-with-task-relevant-combined-local-51jbzix19n.pdf

Object Recognition with Task Relevant Combined Local Features

We propose a real-time image retrieval system which allows users to search target images whose objects are similar to the query in contour, regardless of their sizes and positions appearing in the images. Even in a complicated scene, as long as the object's contour is most salient in the target image, the system is still able to capture it and lists the image in the retrieval results. Therefore, the system has better retrieval rate than existing systems and algorithms. One typical application of the proposed system is to help the computer understand what does the user draw or upload. It is based on the statistical distributions of tags of retrieved images, and the proposed system feeds back some candidate tags related to the query image. Such tags could be used for further retrieval to refine the result list. In addition, the system provides a friendly interactive interface with multiple queries. These queries are from different combinations of tags, a hand-drawn sketch and a natural image, and could help users search images flexibly and conveniently. The system runs on a database of 1.3 million images and could achieve a real-time retrieval speed. The results in the demonstration show excellent retrieval performance of the proposed system.

Guess what you draw: interactive contour-based image retrieval on a million-scale database

The EEG signals recorded during Brain Computer Interfaces (BCIs) are naturally represented by multi-way arrays in spatial, temporal, and frequency domains. In order to effectively extract the underlying components from brain activities which correspond to the specific mental state, we propose the higher-order PLS approach to find the latent variables related to the target labels and then make classification based on latent variables. To this end, the low-dimensional latent space can be optimized by using the higher-order SVD on a cross-product tensor, and the latent variables are considered as shared components between observed data and target output. The EEG signals recorded under the P300-type affective BCI paradigm were used to demonstrate the effectiveness of our new approach.

Higher-order PLS for classification of ERPs with application to BCIs

Semantic segmentation has achieved remarkable progress but remains challenging due to the complex scene, object occlusion, and so on. Some research works have attempted to use extra information such as a depth map to help RGB based semantic segmentation because the depth map could provide complementary geometric cues. However, due to the inaccessibility of depth sensors, depth information is usually unavailable for the test images. In this paper, we leverage only the depth of training images as the privileged information to mine the hard pixels in semantic segmentation, in which depth information is only available for training images but not available for test images. Specifically, we propose a novel Loss Weight Module, which outputs a loss weight map by employing two depth-related measurements of hard pixels: Depth Prediction Error and Depthaware Segmentation Error. The loss weight map is then applied to segmentation loss, with the goal of learning a more robust model by paying more attention to the hard pixels. Besides, we also explore a curriculum learning strategy based on the loss weight map. Meanwhile, to fully mine the hard pixels on different scales, we apply our loss weight module to multi-scale side outputs. Our hard pixels mining method achieves the state-of-the-art results on two benchmark datasets, and even outperforms the methods which need depth input during testing.

Hard Pixel Mining for Depth Privileged Semantic Segmentation.

In this paper a new approach called nonnegative tensor principal component analysis (NTPCA) with sparse constraint is proposed for speech feature extraction. We encode speech as a general higher order tensor in order to extract the robust feature from multiple interrelated feature subspace. First, speech signals are represented by cochleagram based on frequency selectivity at basilar membrane and inner hair cells; Then, a low dimension sparse representation based on tensor structure is extracted by NTPCA for robust speaker modeling. Alternating projection algorithm is used to obtain a stable solution and makes sure the useful information of each subspace in the higher order tensor being preserved. Experiment results demonstrate that our method can increase the recognition accuracy specifically in noise environments.

Liqing Zhang

Papers

Object Recognition with Task Relevant Combined Local Features

Guess what you draw: interactive contour-based image retrieval on a million-scale database

Higher-order PLS for classification of ERPs with application to BCIs

Hard Pixel Mining for Depth Privileged Semantic Segmentation.

Nonnegative Tensor PCA and Application to Speaker Recognition in Noise Environments