There is, I think, something ethereal about i —the square root of minus one. I remember first hearing about it at school. It seemed an odd beast at that time—an intruder hovering on the edge of reality.

Usually familiarity dulls this sense of the bizarre, but in the case of i it was the reverse: over the years the sense of its surreal nature intensified. It seemed that it was impossible to write mathematics that described the real world in …

I and i

다중혈관 관상동맥 환자에서 y-문합을 이용하여 양쪽 내흉동맥만을 사용한 우회술의 조기 성적

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

The measurement of meaning

The PASCAL Visual Object Classes Challenge

The IRIM group is a consortium of French teams supported by the GDR ISIS and working on Multimedia Indexing and Retrieval. This paper describes its participation to the TRECVID 2015 semantic indexing (SIN). Our approach uses a six-stages processing pipelines for computing scores for the likelihood of a video shot to contain a target concept. These scores are then used for producing a ranked list of images or shots that are the most likely to contain the target concept. The pipeline is composed of the following steps: descriptor extraction, descriptor optimization, classification, fusion of descriptor variants, higher-level fusion, and re-ranking. We evaluated a number of different descriptors and tried different fusion strategies. The best IRIM run has a Mean Inferred Average Precision of 0.2947, which ranked it 4th out of 15 participants.

/pdf/irim-at-trecvid-2015-semantic-indexing-2g375kqqxe.pdf

IRIM at TRECVID 2015: Semantic indexing

Traditional deep learning algorithms often fail to generalize when they are tested outside of the domain of training data. Because data distributions can change dynamically in real-life applications once a learned model is deployed, in this paper we are interested in single-source domain generalization (SDG) which aims to develop deep learning algorithms able to generalize from a single training domain where no information about the test domain is available at training time. Firstly, we design two simple MNISTbased SDG benchmarks, namely MNIST Color SDG-MP and MNIST Color SDG-UP, which highlight the two different fundamental SDG issues of increasing difficulties: 1) a class-correlated pattern in the training domain is missing (SDG-MP), or 2) uncorrelated with the class (SDG-UP), in the testing data domain. This is in sharp contrast with the current domain generalization (DG) benchmarks which mix up different correlation and variation factors and thereby make hard to disentangle success or failure factors when benchmarking DG algorithms. We further evaluate several state-of-the-art SDG algorithms through our simple benchmark, namely MNIST Color SDG-MP, and show that the issue SDG-MP is largely unsolved despite of a decade of efforts in developing DG algorithms. Finally, we also propose a partially reversed contrastive loss to encourage intra-class diversity and find less strongly correlated patterns, to deal with SDG-MP and show that the proposed approach is very effective on our MNIST Color SDG-MP benchmark.

/pdf/encouraging-intra-class-diversity-through-a-reverse-3ngs1gesr1.pdf

Encouraging Intra-Class Diversity Through a Reverse Contrastive Loss for Better Single-Source Domain Generalization

We present in this paper a comparison of audio signal codings that we have developed in order to study xiphoidal sounds These sounds are produced by the lower oesophageal sphincter whose dysfonctionnement can be responsible for the gastro-eosophageal reflux phenomenon Our goal is to extract pertinent information from audio signals in order to characterize the pathology of the patients and its intensity, which is generally realized by invasive and traumatic methods such as radiological and manometric investigations Several codings of audio signals are presented and compared They are based on different representations of signals: temporal, frequential and time-scale representations Zipf and inverse Zipf analyses are then performed They allow the extraction of pertinent primitives, not available from standard signal processing methods Finally, a clustering step is realized in order to bring to the fore clusters, linked to pathology characteristics Two clustering methods have been used: a linear discriminant analysis, and a neural network clustering method

Comparison of audio signal codings for Zipf analysis of xiphoidal sounds

In this paper, we present the methods we have proposed and evaluated through the ImageCLEF 2012 Photo Annotation task. More precisely, we have proposed the Histogram of Textual Concepts (HTC) textual feature to capture the relatedness of semantic concepts. In con- trast to term frequency-based text representations mostly used for visual concept detection and annotation, HTC relies on the semantic similar- ity between the user tags and a concept dictionary. Moreover, a Selective Weighted Late Fusion (SWLF) is introduced to combine multiple sources of information which by iteratively selecting and weighting the best fea- tures for each concept at hand to be classied. The results have shown

LIRIS - Imagine at Image CLEF 2012 Photo Annotation Task

We present in this paper an audio signal classification method based on Zipf and inverse Zipf laws. These laws are powerful analysis tools allowing the extraction of information not available by the way of standard methods. The adaptation of Zipf and Inverse Zipf laws to audio signals requires a coding of these signals into literary texts, considered as sequences of patterns. Because these codings are of first importance since they have to bring to the fore relevant information in signals, three types of codings have been developed, depending on the representation of the audio signal it is based on: temporal, frequential and time-scale representations. Once audio signals have been coded, features linked to Zipf and inverse Zipf laws are computed. Finally, the classification step aims at the identification of signals. Four classification methods have been considered as well as a fusion method used to combine these classifiers. In order to evaluate our method, we have analysed medical signals corresponding to swallowing signals containing xiphoidal sounds. The problem is to characterize them according to the gastro-oesophageal reflux pathological state.

Emmanuel Dellandréa

Papers

IRIM at TRECVID 2015: Semantic indexing

Encouraging Intra-Class Diversity Through a Reverse Contrastive Loss for Better Single-Source Domain Generalization

Comparison of audio signal codings for Zipf analysis of xiphoidal sounds

LIRIS - Imagine at Image CLEF 2012 Photo Annotation Task

Inner structure computation for audio signal analysis