scholarly journals CLASSIFICATION UNDER LABEL NOISE BASED ON OUTDATED MAPS

2017 ◽  
Vol IV-1/W1 ◽  
pp. 215-222 ◽  
Author(s):  
A. Maas ◽  
F. Rottensteiner ◽  
C. Heipke
Keyword(s):  
Classical Method ◽  
Image Data ◽  
Training Data ◽  
Label Noise ◽  
Training Technique ◽  
Class Labels ◽  
Noise Tolerant ◽  
Changes Over Time ◽  
Over Time

Supervised classification of remotely sensed images is a classical method for change detection. The task requires training data in the form of image data with known class labels, whose manually generation is time-consuming. If the labels are acquired from the outdated map, the classifier must cope with errors in the training data. These errors, referred to as label noise, typically occur in clusters in object space, because they are caused by land cover changes over time. In this paper we adapt a label noise tolerant training technique for classification, so that the fact that changes affect larger clusters of pixels is considered. We also integrate the existing map into an iterative classification procedure to act as a prior in regions which are likely to contain changes. Our experiments are based on three test areas, using real images with simulated existing databases. Our results show that this method helps to distinguish between real changes over time and false detections caused by misclassification and thus improves the accuracy of the classification results.

scholarly journals USING LABEL NOISE ROBUST LOGISTIC REGRESSION FOR AUTOMATED UPDATING OF TOPOGRAPHIC GEOSPATIAL DATABASES

2016 ◽  
Vol III-7 ◽  
pp. 133-140
Author(s):  
A. Maas ◽  
F. Rottensteiner ◽  
C. Heipke
Keyword(s):  
Classical Method ◽  
Image Data ◽  
Training Data ◽  
Label Noise ◽  
Training Techniques ◽  
Training Technique ◽  
Class Labels ◽  
Noise Robust ◽  
Noise Tolerant ◽  
Geospatial Databases

Supervised classification of remotely sensed images is a classical method to update topographic geospatial databases. The task requires training data in the form of image data with known class labels, whose generation is time-consuming. To avoid this problem one can use the labels from the outdated database for training. As some of these labels may be wrong due to changes in land cover, one has to use training techniques that can cope with wrong class labels in the training data. In this paper we adapt a label noise tolerant training technique to the problem of database updating. No labelled data other than the existing database are necessary. The resulting label image and transition matrix between the labels can help to update the database and to detect changes between the two time epochs. Our experiments are based on different test areas, using real images with simulated existing databases. Our results show that this method can indeed detect changes that would remain undetected if label noise were not considered in training.

scholarly journals USING LABEL NOISE ROBUST LOGISTIC REGRESSION FOR AUTOMATED UPDATING OF TOPOGRAPHIC GEOSPATIAL DATABASES

2016 ◽  
Vol III-7 ◽  
pp. 133-140 ◽  
Author(s):  
A. Maas ◽  
F. Rottensteiner ◽  
C. Heipke
Keyword(s):  
Classical Method ◽  
Image Data ◽  
Training Data ◽  
Label Noise ◽  
Training Techniques ◽  
Training Technique ◽  
Class Labels ◽  
Noise Robust ◽  
Noise Tolerant ◽  
Geospatial Databases

Supervised classification of remotely sensed images is a classical method to update topographic geospatial databases. The task requires training data in the form of image data with known class labels, whose generation is time-consuming. To avoid this problem one can use the labels from the outdated database for training. As some of these labels may be wrong due to changes in land cover, one has to use training techniques that can cope with wrong class labels in the training data. In this paper we adapt a label noise tolerant training technique to the problem of database updating. No labelled data other than the existing database are necessary. The resulting label image and transition matrix between the labels can help to update the database and to detect changes between the two time epochs. Our experiments are based on different test areas, using real images with simulated existing databases. Our results show that this method can indeed detect changes that would remain undetected if label noise were not considered in training.

scholarly journals Incremental Learning for Classification of Unstructured Data Using Extreme Learning Machine

Algorithms ◽  
2018 ◽  
Vol 11 (10) ◽  
pp. 158 ◽  
Author(s):  
Sathya Madhusudhanan ◽  
Suresh Jaganathan ◽  
Jayashree L S
Keyword(s):  
Extreme Learning Machine ◽  
Incremental Learning ◽  
Streaming Data ◽  
Unstructured Data ◽  
Feed Forward Neural Network ◽  
Training Time ◽  
Class Labels ◽  
Learning Machine ◽  
Over Time

Unstructured data are irregular information with no predefined data model. Streaming data which constantly arrives over time is unstructured, and classifying these data is a tedious task as they lack class labels and get accumulated over time. As the data keeps growing, it becomes difficult to train and create a model from scratch each time. Incremental learning, a self-adaptive algorithm uses the previously learned model information, then learns and accommodates new information from the newly arrived data providing a new model, which avoids the retraining. The incrementally learned knowledge helps to classify the unstructured data. In this paper, we propose a framework CUIL (Classification of Unstructured data using Incremental Learning) which clusters the metadata, assigns a label for each cluster and then creates a model using Extreme Learning Machine (ELM), a feed-forward neural network, incrementally for each batch of data arrived. The proposed framework trains the batches separately, reducing the memory resources, training time significantly and is tested with metadata created for the standard image datasets like MNIST, STL-10, CIFAR-10, Caltech101, and Caltech256. Based on the tabulated results, our proposed work proves to show greater accuracy and efficiency.

EFFICIENT CLASSIFICATION OF SCANNED MEDIA USING SPATIAL STATISTICS

2010 ◽  
Vol 24 (06) ◽  
pp. 917-946
Author(s):  
GOZDE UNAL ◽  
GAURAV SHARMA ◽  
REINER ESCHBACH
Keyword(s):  
Spatial Statistics ◽  
High Reliability ◽  
Image Data ◽  
Training Data ◽  
Automated Classification ◽  
Data Set ◽  
Statistical Measures ◽  
Scanned Image ◽  
Scanned Images

Photography, lithography, xerography, and inkjet printing are the dominant technologies for color printing. Images produced on these "different media" are often scanned either for the purpose of copying or creating an electronic representation. For an improved color calibration during scanning, a media identification from the scanned image data is desirable. In this paper, we propose an efficient algorithm for automated classification of input media into four major classes corresponding to photographic, lithographic, xerographic and inkjet. Our technique exploits the strong correlation between the type of input media and the spatial statistics of corresponding images, which are observed in the scanned images. We adopt ideas from spatial statistics literature, and design two spatial statistical measures of dispersion and periodicity, which are computed over spatial point patterns generated from blocks of the scanned image, and whose distributions provide the features for making a decision. We utilize extensive training data and determined well separated decision regions to classify the input media. We validate and tested our classification technique results over an independent extensive data set. The results demonstrate that the proposed method is able to distinguish between the different media with high reliability.

scholarly journals Automatic classification of ovarian cancer types from cytological images using deep convolutional neural networks

2018 ◽  
Vol 38 (3) ◽  
Author(s):  
Miao Wu ◽  
Chuanbo Yan ◽  
Huiqiang Liu ◽  
Qian Liu
Keyword(s):  
Neural Networks ◽  
Ovarian Cancer ◽  
Convolutional Neural Networks ◽  
Image Data ◽  
Training Data ◽  
Serous Carcinoma ◽  
Deep Convolutional Neural Networks ◽  
Ovarian Cancers ◽  
Cancer Types

Ovarian cancer is one of the most common gynecologic malignancies. Accurate classification of ovarian cancer types (serous carcinoma, mucous carcinoma, endometrioid carcinoma, transparent cell carcinoma) is an essential part in the different diagnosis. Computer-aided diagnosis (CADx) can provide useful advice for pathologists to determine the diagnosis correctly. In our study, we employed a Deep Convolutional Neural Networks (DCNN) based on AlexNet to automatically classify the different types of ovarian cancers from cytological images. The DCNN consists of five convolutional layers, three max pooling layers, and two full reconnect layers. Then we trained the model by two group input data separately, one was original image data and the other one was augmented image data including image enhancement and image rotation. The testing results are obtained by the method of 10-fold cross-validation, showing that the accuracy of classification models has been improved from 72.76 to 78.20% by using augmented images as training data. The developed scheme was useful for classifying ovarian cancers from cytological images.

scholarly journals Watermelon Quality Classification Using Weighted K-means Algortihm Based On Feautures YCbCr Color

2019 ◽  
Vol 1 (2) ◽  
Author(s):  
Lalu Zulfikar Muslim ◽  
I Gede Pasek Suta Wijaya ◽  
Fitri Bimantoro
Keyword(s):  
Input Data ◽  
Computing Time ◽  
Image Data ◽  
Training Data ◽  
Business Strategies ◽  
Computational Time ◽  
Training Phase ◽  
Quality Classification ◽  
Two Stages

the classification of fruit quality on a computer using image data is very necessary. In addition, this can also be used in making decisions and policies related to business strategies in the industry. In this research, the quality classification of watermelon was carried out using the Weighted K-Means Algorithm. The classification of watermelon fruit in this study was divided into three groups, namely fresh, medium, and rotten. The classification process in the system created is divided into two stages, namely training and examinations.The data that is input into the system is watermelon image data in YCbCr format. In the training phase, the input data that is processed is image data that has been classified. As for the testing/classification phase, the input data processed is an arbitrary image that has not been classified.The results of the classification with watermelon case studies using the weighted k-means algorithm obtained a conclusion that the greater the amount of training data, the computing time needed for the training and testing process will increase, as well as the level of accuracy, precision and recall of the classification results obtained will also get better. While the greater the number of k values, the computational time needed for the training and testing process will increase, but the level of accuracy, precision, and recall of the results of the classification that gets smaller.

scholarly journals Fast Reaction to Sudden Concept Drift in the Absence of Class Labels

2020 ◽  
Vol 10 (2) ◽  
pp. 606 ◽  
Author(s):  
Osama A. Mahdi ◽  
Eric Pardede ◽  
Nawfal Ali ◽  
Jinli Cao
Keyword(s):  
Data Stream ◽  
Input Data ◽  
Concept Drift ◽  
Web Page ◽  
Target Variable ◽  
Fast Reaction ◽  
True Label ◽  
Class Labels ◽  
Changes Over Time ◽  
Over Time

A data stream can be considered as a sequence of examples that arrive continuously and are potentially unbounded, such as web page visits, sensor readings and call records. One of the serious and challenging problems that appears in a data stream is concept drift. This problem occurs when the relation between the input data and the target variable changes over time. Most existing works make an optimistic assumption that all incoming data are labelled and the class labels are available immediately. However, such an assumption is not always valid. Therefore, a lack of class labels aggravates the problem of concept drift detection. With this motivation, we propose a drift detector that reacts naturally to sudden drifts in the absence of class labels. In a novel way, the proposed detector reacts to concept drift in the absence of class labels, where the true label of an example is not necessary. Instead of monitoring the error estimates, the proposed detector monitors the diversity of a pair of classifiers, where the true label of an example is not necessary to determine whether components disagree. Using several datasets, an experimental evaluation and comparison is conducted against several existing detectors. The experiment results show that the proposed detector can detect drifts with less delay, runtime and memory usage.

Sign in / Sign up

Export Citation Format

Share Document