Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userStreaming convolutional neural networks for end-to-end learning with multi-megapixel images
86
0
0.0
(
0
)
Added by
Hans Pinckaers
Publication date
2019
fields
Informatics Engineering
and research's language is
English
Ask ChatGPT about the research
No Arabic abstract
Due to memory constraints on current hardware, most convolution neural networks (CNN) are trained on sub-megapixel images. For example, most popular datasets in computer vision contain images much less than a megapixel in size (0.09MP for ImageNet and 0.001MP for CIFAR-10). In some domains such as medical imaging, multi-megapixel images are needed to identify the presence of disease accurately. We propose a novel method to directly train convolutional neural networks using any input image size end-to-end. This method exploits the locality of most operations in modern convolutional neural networks by performing the forward and backward pass on smaller tiles of the image. In this work, we show a proof of concept using images of up to 66-megapixels (8192x8192), saving approximately 50GB of memory per image. Using two public challenge datasets, we demonstrate that CNNs can learn to extract relevant information from these large images and benefit from increasing resolution. We improved the area under the receiver-operating characteristic curve from 0.580 (4MP) to 0.706 (66MP) for metastasis detection in breast cancer (CAMELYON17). We also obtained a Spearman correlation metric approaching state-of-the-art performance on the TUPAC16 dataset, from 0.485 (1MP) to 0.570 (16MP). Code to reproduce a subset of the experiments is available at https://github.com/DIAGNijmegen/StreamingCNN.
rate research
Read More
To train deep convolutional neural networks, the input data and the intermediate activations need to be kept in memory to calculate the gradient descent step. Given the limited memory available in the current generation accelerator cards, this limits the maximum dimensions of the input data. We demonstrate a method to train convolutional neural networks holding only parts of the image in memory while giving equivalent results. We quantitatively compare this new way of training convolutional neural networks with conventional training. In addition, as a proof of concept, we train a convolutional neural network with 64 megapixel images, which requires 97% less memory than the conventional approach.
Convolutional Neural Networks (CNNs) are effective models for reducing spectral variations and modeling spectral correlations in acoustic features for automatic speech recognition (ASR). Hybrid speech recognition systems incorporating CNNs with Hidden Markov Models/Gaussian Mixture Models (HMMs/GMMs) have achieved the state-of-the-art in various benchmarks. Meanwhile, Connectionist Temporal Classification (CTC) with Recurrent Neural Networks (RNNs), which is proposed for labeling unsegmented sequences, makes it feasible to train an end-to-end speech recognition system instead of hybrid settings. However, RNNs are computationally expensive and sometimes difficult to train. In this paper, inspired by the advantages of both CNNs and the CTC approach, we propose an end-to-end speech framework for sequence labeling, by combining hierarchical CNNs with CTC directly without recurrent connections. By evaluating the approach on the TIMIT phoneme recognition task, we show that the proposed model is not only computationally efficient, but also competitive with the existing baseline systems. Moreover, we argue that CNNs have the capability to model temporal correlations with appropriate context information.
Recently, the connectionist temporal classification (CTC) model coupled with recurrent (RNN) or convolutional neural networks (CNN), made it easier to train speech recognition systems in an end-to-end fashion. However in real-valued models, time frame components such as mel-filter-bank energies and the cepstral coefficients obtained from them, together with their first and second order derivatives, are processed as individual elements, while a natural alternative is to process such components as composed entities. We propose to group such elements in the form of quaternions and to process these quaternions using the established quaternion algebra. Quaternion numbers and quaternion neural networks have shown their efficiency to process multidimensional inputs as entities, to encode internal dependencies, and to solve many tasks with less learning parameters than real-valued models. This paper proposes to integrate multiple feature views in quaternion-valued convolutional neural network (QCNN), to be used for sequence-to-sequence mapping with the CTC model. Promising results are reported using simple QCNNs in phoneme recognition experiments with the TIMIT corpus. More precisely, QCNNs obtain a lower phoneme error rate (PER) with less learning parameters than a competing model based on real-valued CNNs.
Dialect identification (DID) is a special case of general language identification (LID), but a more challenging problem due to the linguistic similarity between dialects. In this paper, we propose an end-to-end DID system and a Siamese neural network to extract language embeddings. We use both acoustic and linguistic features for the DID task on the Arabic dialectal speech dataset: Multi-Genre Broadcast 3 (MGB-3). The end-to-end DID system was trained using three kinds of acoustic features: Mel-Frequency Cepstral Coefficients (MFCCs), log Mel-scale Filter Bank energies (FBANK) and spectrogram energies. We also investigated a dataset augmentation approach to achieve robust performance with limited data resources. Our linguistic feature research focused on learning similarities and dissimilarities between dialects using the Siamese network, so that we can reduce feature dimensionality as well as improve DID performance. The best system using a single feature set achieves 73% accuracy, while a fusion system using multiple features yields 78% on the MGB-3 dialect test set consisting of 5 dialects. The experimental results indicate that FBANK features achieve slightly better results than MFCCs. Dataset augmentation via speed perturbation appears to add significant robustness to the system. Although the Siamese network with language embeddings did not achieve as good a result as the end-to-end DID system, the two approaches had good synergy when combined together in a fused system.
End-to-end multi-talker speech recognition is an emerging research trend in the speech community due to its vast potential in applications such as conversation and meeting transcriptions. To the best of our knowledge, all existing research works are constrained in the offline scenario. In this work, we propose the Streaming Unmixing and Recognition Transducer (SURT) for end-to-end multi-talker speech recognition. Our model employs the Recurrent Neural Network Transducer (RNN-T) as the backbone that can meet various latency constraints. We study two different model architectures that are based on a speaker-differentiator encoder and a mask encoder respectively. To train this model, we investigate the widely used Permutation Invariant Training (PIT) approach and the Heuristic Error Assignment Training (HEAT) approach. Based on experiments on the publicly available LibriSpeechMix dataset, we show that HEAT can achieve better accuracy compared with PIT, and the SURT model with 150 milliseconds algorithmic latency constraint compares favorably with the offline sequence-to-sequence based baseline model in terms of accuracy.
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
