Multimodal Summarization Research Articles

Research on Multimodal Summarization Based on Sequence to Sequence Model

With the rapid development of information technology in the 21st century, the data on the Internet is growing exponentially, which makes it difficult for users to receive useful information from an endless stream of information sources. Therefore, as a kind of data compression task, automatic summarization technology has gradually become a research hotspot in academia and industry. This paper mainly introduces the multimodal summarization method from sequence to sequence model, then introduces the current multimodal summarization evaluation methods and finally gives some conclusions and thoughts.

UniMS: A Unified Framework for Multimodal Summarization with Knowledge Distillation

With the rapid increase of multimedia data, a large body of literature has emerged to work on multimodal summarization, the majority of which target at refining salient information from textual and image modalities to output a pictorial summary with the most relevant images. Existing methods mostly focus on either extractive or abstractive summarization and rely on the presence and quality of image captions to build image references. We are the first to propose a Unified framework for Multimodal Summarization grounding on BART, UniMS, that integrates extractive and abstractive objectives, as well as selecting the image output. Specially, we adopt knowledge distillation from a vision-language pretrained model to improve image selection, which avoids any requirement on the existence and quality of image captions. Besides, we introduce a visual guided decoder to better integrate textual and visual modalities in guiding abstractive text generation. Results show that our best model achieves a new state-of-the-art result on a large-scale benchmark dataset. The newly involved extractive objective as well as the knowledge distillation technique are proven to bring a noticeable improvement to the multimodal summarization task.

Hierarchical Cross-Modality Semantic Correlation Learning Model for Multimodal Summarization

Multimodal summarization with multimodal output (MSMO) generates a summary with both textual and visual content. Multimodal news report contains heterogeneous contents, which makes MSMO nontrivial. Moreover, it is observed that different modalities of data in the news report correlate hierarchically. Traditional MSMO methods indistinguishably handle different modalities of data by learning a representation for the whole data, which is not directly adaptable to the heterogeneous contents and hierarchical correlation. In this paper, we propose a hierarchical cross-modality semantic correlation learning model (HCSCL) to learn the intra- and inter-modal correlation existing in the multimodal data. HCSCL adopts a graph network to encode the intra-modal correlation. Then, a hierarchical fusion framework is proposed to learn the hierarchical correlation between text and images. Furthermore, we construct a new dataset with relevant image annotation and image object label information to provide the supervision information for the learning procedure. Extensive experiments on the dataset show that HCSCL significantly outperforms the baseline methods in automatic summarization metrics and fine-grained diversity tests.

Inter- and Intra-Modal Contrastive Hybrid Learning Framework for Multimodal Abstractive Summarization.

Internet users are benefiting from technologies of abstractive summarization enabling them to view articles on the internet by reading article summaries only instead of an entire article. However, there are disadvantages to technologies for analyzing articles with texts and images due to the semantic gap between vision and language. These technologies focus more on aggregating features and neglect the heterogeneity of each modality. At the same time, the lack of consideration of intrinsic data properties within each modality and semantic information from cross-modal correlations result in the poor quality of learned representations. Therefore, we propose a novel Inter- and Intra-modal Contrastive Hybrid learning framework which learns to automatically align the multimodal information and maintains the semantic consistency of input/output flows. Moreover, ITCH can be taken as a component to make the model suitable for both supervised and unsupervised learning approaches. Experiments on two public datasets, MMS and MSMO, show that the ITCH performances are better than the current baselines.

Machine learning based fusion algorithm to perform multimodal summarization

Video summarization is a rapidly growing research field which finds its application in various commercial and personal interests due to the massive surge in the amount of video data available in the modern world. The proposed approach uses ResNet-18 for feature extraction and with the help of temporal interest proposals generated for the video sequences, generates a video summary. The ResNet-18 is a convolutional neural network with eighteen layers. The existing methods don’t address the problem of the summary being temporally consistent. The proposed work aims to create a temporally consistent summary. The classification and regression module are implemented to get fixed length inputs of the combined features. After this, the non-maximum suppression algorithm is applied to reduce the redundancy and remove the video segments having poor quality and low confidence-scores. Video summaries are generated using the kernel temporal segmentation (KTS) algorithm which converts a given video segment into video shots. The two standard datasets TVSum and SumMe are used to evaluate the proposed model. It is seen that the F-score obtained on TVSum and SumMe datasets are 56.13 and 45.06 respectively.

트랜스포머기반의 멀티모달 영상자막 생성요약

In this paper, we propose a MASTF methodology, which is a Multimodal Abstractive Summarization based on Transformer. Neural network models applied in the field of generative summaries utilizing conventional multi-modals were techniques utilizing hierarchical attention based on circulating neural networks. Although transformers showed excellent performance in various natural language processing fields, including generative summaries, there were no cases of application in multimodal-based generative summaries. Thus, in this paper, we use transformers to improve the performance of multimodal image subtitle generation summary models. Transformer-based models outperform hierarchical attention-based models by 24.17% on ROUGE-L basis and 10.52% on combining speech and text.

Multi-Tier Attention Network using Term-weighted Question Features for Visual Question Answering

Visual Question Answering (VQA) is a multi-modal challenging task that accepts an image and a natural language question about that image as inputs and desires to find the correct answer. This AI-complete task necessitates the fine-grained joint understanding of the two input modalities. Inspired by the success of attention mechanism in the task of efficient comprehension of visual-language features for VQA, this paper proposes a Multi-Tier Attention Network (MTAN) with the major component being term-weighted question-guided visual attention. Additionally, we introduce a novel Supervised Term Weighting (STW) scheme named ‘qf.obj.cos’ to semantically weight words utilizing the notion of visual object detection. This can be generalized to other vision-language comprehension tasks like image captioning, text-to-image-retrieval, multi-modal summarization etc. In effect, the proposed system allows the generation of more discriminative visual features from the progressive steps of question guided visual attention where question embedding is indeed guided by semantic term weighting. MTAN is quantitatively and qualitatively evaluated on the benchmark DAQUAR dataset and an extensive set of ablations are studied to demonstrate the individual significance of each of the components of the system. Experimental results certify that MTAN performs better than the previous works using the same dataset.

Multimodal Summarization of User-Generated Videos

The exponential growth of user-generated content has increased the need for efficient video summarization schemes. However, most approaches underestimate the power of aural features, while they are designed to work mainly on commercial/professional videos. In this work, we present an approach that uses both aural and visual features in order to create video summaries from user-generated videos. Our approach produces dynamic video summaries, that is, comprising the most “important” parts of the original video, which are arranged so as to preserve their temporal order. We use supervised knowledge from both the aforementioned modalities and train a binary classifier, which learns to recognize the important parts of videos. Moreover, we present a novel user-generated dataset which contains videos from several categories. Every 1 s part of each video from our dataset has been annotated by more than three annotators as being important or not. We evaluate our approach using several classification strategies based on audio, video and fused features. Our experimental results illustrate the potential of our approach.

Graph-based Multimodal Ranking Models for Multimodal Summarization

Multimodal summarization aims to extract the most important information from the multimedia input. It is becoming increasingly popular due to the rapid growth of multimedia data in recent years. There are various researches focusing on different multimodal summarization tasks. However, the existing methods can only generate single-modal output or multimodal output. In addition, most of them need a lot of annotated samples for training, which makes it difficult to be generalized to other tasks or domains. Motivated by this, we propose a unified framework for multimodal summarization that can cover both single-modal output summarization and multimodal output summarization. In our framework, we consider three different scenarios and propose the respective unsupervised graph-based multimodal summarization models without the requirement of any manually annotated document-summary pairs for training: (1) generic multimodal ranking, (2) modal-dominated multimodal ranking, and (3) non-redundant text-image multimodal ranking. Furthermore, an image-text similarity estimation model is introduced to measure the semantic similarity between image and text. Experiments show that our proposed models outperform the single-modal summarization methods on both automatic and human evaluation metrics. Besides, our models can also improve the single-modal summarization with the guidance of the multimedia information. This study can be applied as the benchmark for further study on multimodal summarization task.

LAMS: A Location-aware Approach for Multimodal Summarization (Student Abstract)

Multimodal summarization aims to refine salient information from multiple modalities, among which texts and images are two mostly discussed ones. In recent years, many fantastic works have emerged in this field by modeling image-text interactions; however, they neglect the fact that most of multimodal documents have been elaborately organized by their writers. This means that a critical organized factor has long been short of enough attention, that is, image locations, which may carry illuminating information and imply the key contents of a document. To address this issue, we propose a location-aware approach for multimodal summarization (LAMS) based on Transformer. We investigate image locations for multimodal summarization via a stack of multimodal fusion block, which can formulate the high-order interactions among images and texts. An extensive experimental study on an extended multimodal dataset validates the superior summarization performance of the proposed model.

D-MmT: A concise decoder-only multi-modal transformer for abstractive summarization in videos

Multi-modal abstractive summarization for videos is an emerging task, aiming to integrate multi-modal and multi-source inputs (video, audio transcript) into a compressed textual summary. Although recent multi-encoder-decoder models on this task have shown promising performance, they did not explicitly model interactions of multi-source inputs. While some strategies like co-attention are utilized for modeling this interaction, considering ultra-long sequences and additional decoder in this task, the coupling of multi-modal data from multi-encoders and decoder needs complicated structure and additional parameters. In this paper, we propose a concise Decoder-only Multi-modal Transformer (D-MmT) based on the above observations. Specifically, we cut the encoder structure, and introduce an in-out shared multi-modal decoder to make the multi-source and target fully interact and couple in the shared feature space, reducing the model parameter redundancy. Also, we design a concise cascaded cross-modal interaction (CXMI) module in the multi-modal decoder that generates joint fusion representations and spontaneously establishes a fine-grained intra- and inter- association between multi-modalities. In addition, to make full use of the ultra-long sequence information, we introduce a joint in-out loss to make the input transcript also participate in backpropagation to enhance the contextual feature representation. The experimental results on the How2 dataset show that the proposed model outperforms the current state-of-the-art approach with fewer model parameters. Further analysis and visualization show the effectiveness of our proposed framework.

MBA: A Multimodal Bilinear Attention Model with Residual Connection for Abstractive Multimodal Summarization

The combination of vision and natural language modalities has become an important topic in both computer vision and natural language processing research communities. Multimodal summarization has received unprecedented attention with the rapid growth of multimodal information. This paper proposes MBA which consists of pre-trained feature extractors, text encoder, image encoder, multimodal bilinear attention fusion module, and summary decoder to complete abstractive multimodal summarization task. A residual network is added to the model to enhance the textual modality information and alleviate the modality-bias problem. Experiments show that the model is better than the baseline models and performs better than text summarization methods that ignore visual modality.

Multimodal Summarization with Guidance of Multimodal Reference

Multimodal summarization with multimodal output (MSMO) is to generate a multimodal summary for a multimodal news report, which has been proven to effectively improve users' satisfaction. The existing MSMO methods are trained by the target of text modality, leading to the modality-bias problem that ignores the quality of model-selected image during training. To alleviate this problem, we propose a multimodal objective function with the guidance of multimodal reference to use the loss from the summary generation and the image selection. Due to the lack of multimodal reference data, we present two strategies, i.e., ROUGE-ranking and Order-ranking, to construct the multimodal reference by extending the text reference. Meanwhile, to better evaluate multimodal outputs, we propose a novel evaluation metric based on joint multimodal representation, projecting the model output and multimodal reference into a joint semantic space during evaluation. Experimental results have shown that our proposed model achieves the new state-of-the-art on both automatic and manual evaluation metrics. Besides, our proposed evaluation method can effectively improve the correlation with human judgments.

Aspect-Aware Multimodal Summarization for Chinese E-Commerce Products

We present an abstractive summarization system that produces summary for Chinese e-commerce products. This task is more challenging than general text summarization. First, the appearance of a product typically plays a significant role in customers' decisions to buy the product or not, which requires that the summarization model effectively use the visual information of the product. Furthermore, different products have remarkable features in various aspects, such as “energy efficiency” and “large capacity” for refrigerators. Meanwhile, different customers may care about different aspects. Thus, the summarizer needs to capture the most attractive aspects of a product that resonate with potential purchasers. We propose an aspect-aware multimodal summarization model that can effectively incorporate the visual information and also determine the most salient aspects of a product. We construct a large-scale Chinese e-commerce product summarization dataset that contains approximately 1.4 million manually created product summaries that are paired with detailed product information, including an image, a title, and other textual descriptions for each product. The experimental results on this dataset demonstrate that our models significantly outperform the comparative methods in terms of both the ROUGE score and manual evaluations.

Multi-Modal Deep Analysis for Multimedia

With the rapid development of Internet and multimedia services in the past decade, a huge amount of user-generated and service provider-generated multimedia data become available. These data are heterogeneous and multi-modal in nature, imposing great challenges for processing and analyzing them. Multi-modal data consist of a mixture of various types of data from different modalities such as texts, images, videos, audios etc. In this article, we present a deep and comprehensive overview for multi-modal analysis in multimedia. We introduce two scientific research problems, data-driven correlational representation and knowledge-guided fusion for multimedia analysis. To address the two scientific problems, we investigate them from the following aspects: 1) multi-modal correlational representation: multi-modal fusion of data across different modalities, and 2) multi-modal data and knowledge fusion: multi-modal fusion of data with domain knowledge. More specifically, on data-driven correlational representation, we highlight three important categories of methods, such as multi-modal deep representation, multi-modal transfer learning, and multi-modal hashing. On knowledge-guided fusion, we discuss the approaches for fusing knowledge with data and four exemplar applications that require various kinds of domain knowledge, including multi-modal visual question answering, multi-modal video summarization, multi-modal visual pattern mining and multi-modal recommendation. Finally, we bring forward our insights and future research directions.

Text summarization using multiobjective optimization

In this article, I have briefly described one of my research areas namely text summarization. The problem definition, state of the art techniques and some preliminary results are presented. There are different types of summarization systems available like document summarization, microblog summarization, multimodal summarization, figure-assisted summarization. These different areas are described briefly in the current article.

Read, Watch, Listen, and Summarize: Multi-Modal Summarization for Asynchronous Text, Image, Audio and Video

Automatic text summarization is a fundamental natural language processing (NLP) application that aims to condense a source text into a shorter version. The rapid increase in multimedia data transmission over the Internet necessitates multi-modal summarization (MMS) from asynchronous collections of text, image, audio, and video. In this work, we propose an extractive MMS method that unites the techniques of NLP, speech processing, and computer vision to explore the rich information contained in multi-modal data and to improve the quality of multimedia news summarization. The key idea is to bridge the semantic gaps between multi-modal content. Audio and visual are main modalities in the video. For audio information, we design an approach to selectively use its transcription and to infer the salience of the transcription with audio signals. For visual information, we learn the joint representations of text and images using a neural network. Then, we capture the coverage of the generated summary for important visual information through text-image matching or multi-modal topic modeling. Finally, all the multi-modal aspects are considered to generate a textual summary by maximizing the salience, non-redundancy, readability, and coverage through the budgeted optimization of submodular functions. We further introduce a publicly available MMS corpus in English and Chinese.1 The experimental results obtained on our dataset demonstrate that our methods based on image matching and image topic framework outperform other competitive baseline methods.

Extractive summarization of documents with images based on multi-modal RNN

Rapid growth of multi-modal documents containing images on the Internet expresses strong demand on multi-modal summarization. The challenge is to create a computing method that can uniformly process text and image. Deep learning provides basic models for meeting this challenge. This paper treats extractive multi-modal summarization as a classification problem and proposes a sentence–image classification method based on the multi-modal RNN model. Our method encodes words and sentences with the hierarchical RNN models and encodes the ordered image set with the CNN model and the RNN model, and then calculates the selection probability of sentences and the sentence–image alignment probability through a logistic classifier taking text coverage, text redundancy, image set coverage, and image set redundancy as features. Two methods are proposed to compute the image set redundancy feature by combining the important scores of sentences and the hidden sentence–image alignment. Experiments on the extended DailyMail corpora constructed by collecting images and captions from the Web show that our method outperforms 11 baseline text summarization methods and that adopting the two image-related features in the classification method can improve text summarization. Our method is able to mine the hidden sentence–image alignments and to create informative well-aligned multi-modal summaries.

You Are What You Eat: Exploring Rich Recipe Information for Cross-Region Food Analysis

Cuisine is a style of cooking and usually associated with a specific geographic region. Recipes from different cuisines shared on the web are an indicator of culinary cultures in different countries. Therefore, analysis of these recipes can lead to deep understanding of food from the cultural perspective. In this paper, we perform the first cross-region recipe analysis by jointly using the recipe ingredients, food images, and attributes such as the cuisine and course (e.g., main dish and dessert). For that solution, we propose a culinary culture analysis framework to discover the topics of ingredient bases and visualize them to enable various applications. We first propose a probabilistic topic model to discover cuisine-course specific topics. The manifold ranking method is then utilized to incorporate deep visual features to retrieve food images for topic visualization. At last, we applied the topic modeling and visualization method for three applications: 1) multimodal cuisine summarization with both recipe ingredients and images, 2) cuisine-course pattern analysis including topic-specific cuisine distribution and cuisine-specific course distribution of topics, and 3) cuisine recommendation for both cuisine-oriented and ingredient-oriented queries. Through these three applications, we can analyze the culinary cultures at both macro and micro levels. We conduct the experiment on a recipe database Yummly-66K with 66,615 recipes from 10 cuisines in Yummly. Qualitative and quantitative evaluation results have validated the effectiveness of topic modeling and visualization, and demonstrated the advantage of the framework in utilizing rich recipe information to analyze and interpret the culinary cultures from different regions.

Multimedia Pivot Tables for Multimedia Analytics on Image Collections

We propose a multimedia analytics solution for getting insight into image collections by extending the powerful analytic capabilities of pivot tables, found in the ubiquitous spreadsheets, to multimedia. We formalize the concept of multimedia pivot tables and give design rules and methods for the multimodal summarization, structuring, and browsing of the collection based on these tables, all optimized to support an analyst in getting structural and conclusive insights. Our proposed solution provides truly interactive analytics on the visual content of image collections through concept detection results, as well as tags, geolocation, time, and other metadata. We have performed user experiments with novice users on a dataset from Flickr to improve the initial design and with expert users in marketing and multimedia analysis on two domain-specific datasets collected from Instagram. The results show that analysts are indeed capable of deriving structural and conclusive insights using the proposed multimedia analytics solution. On our website, videos of the system in action are available. 1