Xin Tang

Hu Hu, Chao-Han Huck Yang, Xianjun Xia et al.

In this technical report, we present a joint effort of four groups, namely GT, USTC, Tencent, and UKE, to tackle Task 1 - Acoustic Scene Classification (ASC) in the DCASE 2020 Challenge. Task 1 comprises two different sub-tasks: (i) Task 1a focuses on ASC of audio signals recorded with multiple (real and simulated) devices into ten different fine-grained classes, and (ii) Task 1b concerns with classification of data into three higher-level classes using low-complexity solutions. For Task 1a, we propose a novel two-stage ASC system leveraging upon ad-hoc score combination of two convolutional neural networks (CNNs), classifying the acoustic input according to three classes, and then ten classes, respectively. Four different CNN-based architectures are explored to implement the two-stage classifiers, and several data augmentation techniques are also investigated. For Task 1b, we leverage upon a quantization method to reduce the complexity of two of our top-accuracy three-classes CNN-based architectures. On Task 1a development data set, an ASC accuracy of 76.9\% is attained using our best single classifier and data augmentation. An accuracy of 81.9\% is then attained by a final model fusion of our two-stage ASC classifiers. On Task 1b development data set, we achieve an accuracy of 96.7\% with a model size smaller than 500KB. Code is available: https://github.com/MihawkHu/DCASE2020_task1.

Xin Tang, Yongquan Lai, Ying Liu et al.

Modeling semantic information is helpful for scene text recognition. In this work, we propose to model semantic and visual information jointly with a Visual-Semantic Transformer (VST). The VST first explicitly extracts primary semantic information from visual feature maps with a transformer module and a primary visual-semantic alignment module. The semantic information is then joined with the visual feature maps (viewed as a sequence) to form a pseudo multi-domain sequence combining visual and semantic information, which is subsequently fed into an transformer-based interaction module to enable learning of interactions between visual and semantic features. In this way, the visual features can be enhanced by the semantic information and vice versus. The enhanced version of visual features are further decoded by a secondary visual-semantic alignment module which shares weights with the primary one. Finally, the decoded visual features and the enhanced semantic features are jointly processed by the third transformer module obtaining the final text prediction. Experiments on seven public benchmarks including regular/ irregular text recognition datasets verifies the effectiveness our proposed model, reaching state of the art on four of the seven benchmarks.

Hu Hu, Chao-Han Huck Yang, Xianjun Xia et al.

To improve device robustness, a highly desirable key feature of a competitive data-driven acoustic scene classification (ASC) system, a novel two-stage system based on fully convolutional neural networks (CNNs) is proposed. Our two-stage system leverages on an ad-hoc score combination based on two CNN classifiers: (i) the first CNN classifies acoustic inputs into one of three broad classes, and (ii) the second CNN classifies the same inputs into one of ten finer-grained classes. Three different CNN architectures are explored to implement the two-stage classifiers, and a frequency sub-sampling scheme is investigated. Moreover, novel data augmentation schemes for ASC are also investigated. Evaluated on DCASE 2020 Task 1a, our results show that the proposed ASC system attains a state-of-the-art accuracy on the development set, where our best system, a two-stage fusion of CNN ensembles, delivers a 81.9% average accuracy among multi-device test data, and it obtains a significant improvement on unseen devices. Finally, neural saliency analysis with class activation mapping (CAM) gives new insights on the patterns learnt by our models.

Bingcong Li, Xin Tang, Xianbiao Qi et al.

Recently, inspired by Transformer, self-attention-based scene text recognition approaches have achieved outstanding performance. However, we find that the size of model expands rapidly with the lexicon increasing. Specifically, the number of parameters for softmax classification layer and output embedding layer are proportional to the vocabulary size. It hinders the development of a lightweight text recognition model especially applied for Chinese and multiple languages. Thus, we propose a lightweight scene text recognition model named Hamming OCR. In this model, a novel Hamming classifier, which adopts locality sensitive hashing (LSH) algorithm to encode each character, is proposed to replace the softmax regression and the generated LSH code is directly employed to replace the output embedding. We also present a simplified transformer decoder to reduce the number of parameters by removing the feed-forward network and using cross-layer parameter sharing technique. Compared with traditional methods, the number of parameters in both classification and embedding layers is independent on the size of vocabulary, which significantly reduces the storage requirement without loss of accuracy. Experimental results on several datasets, including four public benchmaks and a Chinese text dataset synthesized by SynthText with more than 20,000 characters, shows that Hamming OCR achieves competitive results.

Jiaxin Cai, Xin Tang

Action recognition has attracted increasing attention from RGB input in computer vision partially due to potential applications on somatic simulation and statistics of sport such as virtual tennis game and tennis techniques and tactics analysis by video. Recently, deep learning based methods have achieved promising performance for action recognition. In this paper, we propose weighted Long Short-Term Memory adopted with convolutional neural network representations for three dimensional tennis shots recognition. First, the local two-dimensional convolutional neural network spatial representations are extracted from each video frame individually using a pre-trained Inception network. Then, a weighted Long Short-Term Memory decoder is introduced to take the output state at time t and the historical embedding feature at time t-1 to generate feature vector using a score weighting scheme. Finally, we use the adopted CNN and weighted LSTM to map the original visual features into a vector space to generate the spatial-temporal semantical description of visual sequences and classify the action video content. Experiments on the benchmark demonstrate that our method using only simple raw RGB video can achieve better performance than the state-of-the-art baselines for tennis shot recognition.

Xin Jie Tang, Yong Haur Tay, Nordahlia Abdullah Siam et al.

Wood Identification has never been more important to serve the purpose of global forest species protection and timber regulation. Macroscopic level wood identification practiced by wood anatomists can identify wood up to genus level. This is sufficient to serve as a frontline identification to fight against illegal wood logging and timber trade for law enforcement authority. However, frontline enforcement official may lack of the accuracy and confidence of a well trained wood anatomist. Hence, computer assisted method such as machine vision methods are developed to do rapid field identification for law enforcement official. In this paper, we proposed a rapid and robust macroscopic wood identification system using machine vision method with off-the-shelf smartphone and retrofitted macro-lens. Our system is cost effective, easily accessible, fast and scalable at the same time provides human-level accuracy on identification. Camera-enabled smartphone with Internet connectivity coupled with a macro-lens provides a simple and effective digital acquisition of macroscopic wood images which are essential for macroscopic wood identification. The images are immediately streamed to a cloud server via Internet connection for identification which are done within seconds.

Jia-xin Cai, Xin Tang, Lifang Zhang et al.

In this paper, a discriminative two-phase dictionary learning framework is proposed for classifying human action by sparse shape representations, in which the first-phase dictionary is learned on the selected discriminative frames and the second-phase dictionary is built for recognition using reconstruction errors of the first-phase dictionary as input features. We propose a "zeroth class" trick for detecting undiscriminating frames of the test video and eliminating them before voting on the action categories. Experimental results on benchmarks demonstrate the effectiveness of our method.