Scene Text Recognition (STR), the task of recognizing text against complex image backgrounds, is an active area of research. Current state-of-the-art (SOTA) methods still struggle to recognize text written in arbitrary shapes. In this paper, we introduce a novel architecture for STR, named Selective Context ATtentional Text Recognizer (SCATTER). SCATTER utilizes a stacked block architecture with intermediate supervision during training, that paves the way to successfully train a deep BiLSTM encoder, thus improving the encoding of contextual dependencies. Decoding is done using a two-step 1D attention mechanism. The first attention step re-weights visual features from a CNN backbone together with contextual features computed by a BiLSTM layer. The second attention step, similar to previous papers, treats the features as a sequence and attends to the intra-sequence relationships. Experiments show that the proposed approach surpasses SOTA performance on irregular text recognition benchmarks by 3.7\% on average.
We present a cross-modal Transformer-based framework, which jointly encodes video data and text labels for zero-shot action recognition (ZSAR). Our model employs a conceptually new pipeline by which visual representations are learned in conjunction with visual-semantic associations in an end-to-end manner. The model design provides a natural mechanism for visual and semantic representations to be learned in a shared knowledge space, whereby it encourages the learned visual embedding to be discriminative and more semantically consistent. In zero-shot inference, we devise a simple semantic transfer scheme that embeds semantic relatedness information between seen and unseen classes to composite unseen visual prototypes. Accordingly, the discriminative features in the visual structure could be preserved and exploited to alleviate the typical zero-shot issues of information loss, semantic gap, and the hubness problem. Under a rigorous zero-shot setting of not pre-training on additional datasets, the experiment results show our model considerably improves upon the state of the arts in ZSAR, reaching encouraging top-1 accuracy on UCF101, HMDB51, and ActivityNet benchmark datasets. Code will be made available.
Pretrained Transformer-based language models (LMs) display remarkable natural language generation capabilities. With their immense potential, controlling text generation of such LMs is getting attention. While there are studies that seek to control high-level attributes (such as sentiment and topic) of generated text, there is still a lack of more precise control over its content at the word- and phrase-level. Here, we propose Content-Conditioner (CoCon) to control an LM's output text with a target content, at a fine-grained level. In our self-supervised approach, the CoCon block learns to help the LM complete a partially-observed text sequence by conditioning with content inputs that are withheld from the LM. Through experiments, we show that CoCon can naturally incorporate target content into generated texts and control high-level text attributes in a zero-shot manner.
Text-to-image synthesis is the task of generating images from text descriptions. Image generation, by itself, is a challenging task. When we combine image generation and text, we bring complexity to a new level: we need to combine data from two different modalities. Most of recent works in text-to-image synthesis follow a similar approach when it comes to neural architectures. Due to aforementioned difficulties, plus the inherent difficulty of training GANs at high resolutions, most methods have adopted a multi-stage training strategy. In this paper we shift the architectural paradigm currently used in text-to-image methods and show that an effective neural architecture can achieve state-of-the-art performance using a single stage training with a single generator and a single discriminator. We do so by applying deep residual networks along with a novel sentence interpolation strategy that enables learning a smooth conditional space. Finally, our work points a new direction for text-to-image research, which has not experimented with novel neural architectures recently.
Synthetic data used for scene text detection and recognition tasks have proven effective. However, there are still two problems: First, the color schemes used for text coloring in the existing methods are relatively fixed color key-value pairs learned from real datasets. The dirty data in real datasets may cause the problem that the colors of text and background are too similar to be distinguished from each other. Second, the generated texts are uniformly limited to the same depth of a picture, while there are special cases in the real world that text may appear across depths. To address these problems, in this paper we design a novel method to generate color schemes, which are consistent with the characteristics of human eyes to observe things. The advantages of our method are as follows: (1) overcomes the color confusion problem between text and background caused by dirty data; (2) the texts generated are allowed to appear in most locations of any image, even across depths; (3) avoids analyzing the depth of background, such that the performance of our method exceeds the state-of-the-art methods; (4) the speed of generating images is fast, nearly one picture generated per three milliseconds. The effectiveness of our method is verified on several public datasets.
A scene text spotter is composed of text detection and recognition modules. Many studies have been conducted to unify these modules into an end-to-end trainable model to achieve better performance. A typical architecture places detection and recognition modules into separate branches, and a RoI pooling is commonly used to let the branches share a visual feature. However, there still exists a chance of establishing a more complimentary connection between the modules when adopting recognizer that uses attention-based decoder and detector that represents spatial information of the character regions. This is possible since the two modules share a common sub-task which is to find the location of the character regions. Based on the insight, we construct a tightly coupled single pipeline model. This architecture is formed by utilizing detection outputs in the recognizer and propagating the recognition loss through the detection stage. The use of character score map helps the recognizer attend better to the character center points, and the recognition loss propagation to the detector module enhances the localization of the character regions. Also, a strengthened sharing stage allows feature rectification and boundary localization of arbitrary-shaped text regions. Extensive experiments demonstrate state-of-the-art performance in publicly available straight and curved benchmark dataset.
Entity linking (EL) is the task of linking entity mentions in a document to referent entities in a knowledge base (KB). Many previous studies focus on Wikipedia-derived KBs. There is little work on EL over Wikidata, even though it is the most extensive crowdsourced KB. The scale of Wikidata can open up many new real-world applications, but its massive number of entities also makes EL challenging. To effectively narrow down the search space, we propose a novel candidate retrieval paradigm based on entity profiling. Wikidata entities and their textual fields are first indexed into a text search engine (e.g., Elasticsearch). During inference, given a mention and its context, we use a sequence-to-sequence (seq2seq) model to generate the profile of the target entity, which consists of its title and description. We use the profile to query the indexed search engine to retrieve candidate entities. Our approach complements the traditional approach of using a Wikipedia anchor-text dictionary, enabling us to further design a highly effective hybrid method for candidate retrieval. Combined with a simple cross-attention reranker, our complete EL framework achieves state-of-the-art results on three Wikidata-based datasets and strong performance on TACKBP-2010.
Over the last few years, Text classification is one of the fundamental tasks in natural language processing (NLP) in which the objective is to categorize text documents into one of the predefined classes. The news is full of our life. Therefore, news headlines classification is a crucial task to connect users with the right news. The news headline classification is a kind of text classification, which can be generally divided into three mainly parts: feature extraction, classifier selection, and evaluations. In this article, we use the dataset, containing news over a period of eighteen years provided by Kaggle platform to classify news headlines. We choose TF-IDF to extract features and neural network as the classifier, while the evaluation metrics is accuracy. From the experiment result, it is obvious that our NN model has the best performance among these models in the metrics of accuracy. The higher the accuracy is, the better performance the model will gain. Our NN model owns the accuracy 0.8622, which is highest accuracy among these four models. And it is 0.0134, 0.033, 0.080 higher than its of other models.
Sentiment Analysis typically refers to using natural language processing, text analysis and computational linguistics to extract affect and emotion based information from text data. Our work explores how we can effectively use deep neural networks in transfer learning and joint dual input learning settings to effectively classify sentiments and detect hate speech in Hindi and Bengali data. We start by training Word2Vec word embeddings for Hindi \textbf{HASOC dataset} and Bengali hate speech and then train LSTM and subsequently, employ parameter sharing based transfer learning to Bengali sentiment classifiers by reusing and fine-tuning the trained weights of Hindi classifiers with both classifier being used as baseline in our study. Finally, we use BiLSTM with self attention in joint dual input learning setting where we train a single neural network on Hindi and Bengali dataset simultaneously using their respective embeddings.
In this paper, we present LaTeX-Numeric - a high-precision fully-automated scalable framework for extracting E-commerce numeric attributes from product text like product description. Most of the past work on attribute extraction is not scalable as they rely on manually curated training data, either with or without the use of active learning. We rely on distant supervision for training data generation, removing dependency on manual labels. One issue with distant supervision is that it leads to incomplete training annotation due to missing attribute values while matching. We propose a multi-task learning architecture to deal with missing labels in the training data, leading to F1 improvement of 9.2% for numeric attributes over single-task architecture. While multi-task architecture benefits both numeric and non-numeric attributes, we present automated techniques to further improve the numeric attributes extraction models. Numeric attributes require a list of units (or aliases) for better matching with distant supervision. We propose an automated algorithm for alias creation using product text and attribute values, leading to a 20.2% F1 improvement. Extensive experiments on real world dataset for 20 numeric attributes across 5 product categories and 3 English marketplaces show that LaTeX-Numeric achieves a high F1-score, without any manual intervention, making it suitable for practical applications. Finally, we show that the improvements are language-agnostic and LaTeX-Numeric achieves 13.9% F1 improvement for 3 Romance languages.