Deep Learning has implemented a wide range of applications and has become increasingly popular in recent years. The goal of multimodal deep learning is to create models that can process and link information using various modalities. Despite the extensive development made for unimodal learning, it still cannot cover all the aspects of human learning. Multimodal learning helps to understand and analyze better when various senses are engaged in the processing of information. This paper focuses on multiple types of modalities, i.e., image, video, text, audio, body gestures, facial expressions, and physiological signals. Detailed analysis of past and current baseline approaches and an in-depth study of recent advancements in multimodal deep learning applications has been provided. A fine-grained taxonomy of various multimodal deep learning applications is proposed, elaborating on different applications in more depth. Architectures and datasets used in these applications are also discussed, along with their evaluation metrics. Last, main issues are highlighted separately for each domain along with their possible future research directions.
Billions of photos are uploaded to the web daily through various types of social networks. Some of these images receive millions of views and become popular, whereas others remain completely unnoticed. This raises the problem of predicting image popularity on social media. The popularity of an image can be affected by several factors, such as visual content, aesthetic quality, user, post metadata, and time. Thus, considering all these factors is essential for accurately predicting image popularity. In addition, the efficiency of the predictive model also plays a crucial role. In this study, motivated by multimodal learning, which uses information from various modalities, and the current success of convolutional neural networks (CNNs) in various fields, we propose a deep learning model, called visual-social convolutional neural network (VSCNN), which predicts the popularity of a posted image by incorporating various types of visual and social features into a unified network model. VSCNN first learns to extract high-level representations from the input visual and social features by utilizing two individual CNNs. The outputs of these two networks are then fused into a joint network to estimate the popularity score in the output layer. We assess the performance of the proposed method by conducting extensive experiments on a dataset of approximately 432K images posted on Flickr. The simulation results demonstrate that the proposed VSCNN model significantly outperforms state-of-the-art models, with a relative improvement of greater than 2.33%, 7.59%, and 14.16% in terms of Spearman's Rho, mean absolute error, and mean squared error, respectively.
We propose a novel, efficient, modular and scalable framework for content based visual media retrieval systems by leveraging the power of Deep Learning which is flexible to work both for images and videos conjointly and we also introduce an efficient comparison and filtering metric for retrieval. We put forward our findings from critical performance tests comparing our method to the predominant conventional approach to demonstrate the feasibility and efficiency of the proposed solution with best practices, possible improvements that may further augment the ability of retrieval architectures.
Explaining the decision of a multi-modal decision-maker requires to determine the evidence from both modalities. Recent advances in XAI provide explanations for models trained on still images. However, when it comes to modeling multiple sensory modalities in a dynamic world, it remains underexplored how to demystify the mysterious dynamics of a complex multi-modal model. In this work, we take a crucial step forward and explore learnable explanations for audio-visual recognition. Specifically, we propose a novel space-time attention network that uncovers the synergistic dynamics of audio and visual data over both space and time. Our model is capable of predicting the audio-visual video events, while justifying its decision by localizing where the relevant visual cues appear, and when the predicted sounds occur in videos. We benchmark our model on three audio-visual video event datasets, comparing extensively to multiple recent multi-modal representation learners and intrinsic explanation models. Experimental results demonstrate the clear superior performance of our model over the existing methods on audio-visual video event recognition. Moreover, we conduct an in-depth study to analyze the explainability of our model based on robustness analysis via perturbation tests and pointing games using human annotations.
Memories are the tethering threads that tie us to the world, and memorability is the measure of their tensile strength. The threads of memory are spun from fibres of many modalities, obscuring the contribution of a single fibre to a thread's overall tensile strength. Unfurling these fibres is the key to understanding the nature of their interaction, and how we can ultimately create more meaningful media content. In this paper, we examine the influence of audio on video recognition memorability, finding evidence to suggest that it can facilitate overall video recognition memorability rich in high-level (gestalt) audio features. We introduce a novel multimodal deep learning-based late-fusion system that uses audio gestalt to estimate the influence of a given video's audio on its overall short-term recognition memorability, and selectively leverages audio features to make a prediction accordingly. We benchmark our audio gestalt based system on the Memento10k short-term video memorability dataset, achieving top-2 state-of-the-art results.
The assignment of importance scores to particular frames or (short) segments in a video is crucial for summarization, but also a difficult task. Previous work utilizes only one source of visual features. In this paper, we suggest a novel model architecture that combines three feature sets for visual content and motion to predict importance scores. The proposed architecture utilizes an attention mechanism before fusing motion features and features representing the (static) visual content, i.e., derived from an image classification model. Comprehensive experimental evaluations are reported for two well-known datasets, SumMe and TVSum. In this context, we identify methodological issues on how previous work used these benchmark datasets, and present a fair evaluation scheme with appropriate data splits that can be used in future work. When using static and motion features with parallel attention mechanism, we improve state-of-the-art results for SumMe, while being on par with the state of the art for the other dataset.
Having access to multi-modal cues (e.g. vision and audio) empowers some cognitive tasks to be done faster compared to learning from a single modality. In this work, we propose to transfer knowledge across heterogeneous modalities, even though these data modalities may not be semantically correlated. Rather than directly aligning the representations of different modalities, we compose audio, image, and video representations across modalities to uncover richer multi-modal knowledge. Our main idea is to learn a compositional embedding that closes the cross-modal semantic gap and captures the task-relevant semantics, which facilitates pulling together representations across modalities by compositional contrastive learning. We establish a new, comprehensive multi-modal distillation benchmark on three video datasets: UCF101, ActivityNet, and VGGSound. Moreover, we demonstrate that our model significantly outperforms a variety of existing knowledge distillation methods in transferring audio-visual knowledge to improve video representation learning. Code is released here: https://github.com/yanbeic/CCL.
With recent advances in autonomous driving, Voice Control Systems have become increasingly adopted as human-vehicle interaction methods. This technology enables drivers to use voice commands to control the vehicle and will be soon available in Advanced Driver Assistance Systems (ADAS). Prior work has shown that Siri, Alexa and Cortana, are highly vulnerable to inaudible command attacks. This could be extended to ADAS in real-world applications and such inaudible command threat is difficult to detect due to microphone nonlinearities. In this paper, we aim to develop a more practical solution by using camera views to defend against inaudible command attacks where ADAS are capable of detecting their environment via multi-sensors. To this end, we propose a novel multimodal deep learning classification system to defend against inaudible command attacks. Our experimental results confirm the feasibility of the proposed defense methods and the best classification accuracy reaches 89.2%. Code is available at https://github.com/ITSEG-MQ/Sensor-Fusion-Against-VoiceCommand-Attacks.
The exponential rise of online social media has enabled the creation, distribution, and consumption of information at an unprecedented rate. However, it has also led to the burgeoning of various forms of online abuse. Increasing cases of online antisemitism have become one of the major concerns because of its socio-political consequences. Unlike other major forms of online abuse like racism, sexism, etc., online antisemitism has not been studied much from a machine learning perspective. To the best of our knowledge, we present the first work in the direction of automated multimodal detection of online antisemitism. The task poses multiple challenges that include extracting signals across multiple modalities, contextual references, and handling multiple aspects of antisemitism. Unfortunately, there does not exist any publicly available benchmark corpus for this critical task. Hence, we collect and label two datasets with 3,102 and 3,509 social media posts from Twitter and Gab respectively. Further, we present a multimodal deep learning system that detects the presence of antisemitic content and its specific antisemitism category using text and images from posts. We perform an extensive set of experiments on the two datasets to evaluate the efficacy of the proposed system. Finally, we also present a qualitative analysis of our study.
We introduce a discriminative multimodal descriptor based on a pair of sensor readings: a point cloud from a LiDAR and an image from an RGB camera. Our descriptor, named MinkLoc++, can be used for place recognition, re-localization and loop closure purposes in robotics or autonomous vehicles applications. We use late fusion approach, where each modality is processed separately and fused in the final part of the processing pipeline. The proposed method achieves state-of-the-art performance on standard place recognition benchmarks. We also identify dominating modality problem when training a multimodal descriptor. The problem manifests itself when the network focuses on a modality with a larger overfit to the training data. This drives the loss down during the training but leads to suboptimal performance on the evaluation set. In this work we describe how to detect and mitigate such risk when using a deep metric learning approach to train a multimodal neural network. Our code is publicly available on the project website: https://github.com/jac99/MinkLocMultimodal.
Malicious applications (particularly those targeting the Android platform) pose a serious threat to developers and end-users. Numerous research efforts have been devoted to developing effective approaches to defend against Android malware. However, given the explosive growth of Android malware and the continuous advancement of malicious evasion technologies like obfuscation and reflection, Android malware defense approaches based on manual rules or traditional machine learning may not be effective. In recent years, a dominant research field called deep learning (DL), which provides a powerful feature abstraction ability, has demonstrated a compelling and promising performance in a variety of areas, like natural language processing and computer vision. To this end, employing deep learning techniques to thwart Android malware attacks has recently garnered considerable research attention. Yet, no systematic literature review focusing on deep learning approaches for Android Malware defenses exists. In this paper, we conducted a systematic literature review to search and analyze how deep learning approaches have been applied in the context of malware defenses in the Android environment. As a result, a total of 132 studies covering the period 2014-2021 were identified. Our investigation reveals that, while the majority of these sources mainly consider DL-based on Android malware detection, 53 primary studies (40.1 percent) design defense approaches based on other scenarios. This review also discusses research trends, research focuses, challenges, and future research directions in DL-based Android malware defenses.
Can a computer determine a piano player's skill level? Is it preferable to base this assessment on visual analysis of the player's performance or should we trust our ears over our eyes? Since current CNNs have difficulty processing long video videos, how can shorter clips be sampled to best reflect the players skill level? In this work, we collect and release a first-of-its-kind dataset for multimodal skill assessment focusing on assessing piano player's skill level, answer the asked questions, initiate work in automated evaluation of piano playing skills and provide baselines for future work. Dataset is available from: https://github.com/ParitoshParmar/Piano-Skills-Assessment.
The modern digital world is becoming more and more multimodal. Looking on the internet, images are often associated with the text, so classification problems with these two modalities are very common. In this paper, we examine multimodal classification using textual information and visual representations of the same concept. We investigate two main basic methods to perform multimodal fusion and adapt them with stacking techniques to better handle this type of problem. Here, we use UPMC Food-101, which is a difficult and noisy multimodal dataset that well represents this category of multimodal problems. Our results show that the proposed early fusion technique combined with a stacking-based approach exceeds the state of the art on the dataset used.
Fine-grained image classification is a challenging task due to the presence of hierarchical coarse-to-fine-grained distribution in the dataset. Generally, parts are used to discriminate various objects in fine-grained datasets, however, not all parts are beneficial and indispensable. In recent years, natural language descriptions are used to obtain information on discriminative parts of the object. This paper leverages on natural language description and proposes a strategy for learning the joint representation of natural language description and images using a two-branch network with multiple layers to improve the fine-grained classification task. Extensive experiments show that our approach gains significant improvements in accuracy for the fine-grained image classification task. Furthermore, our method achieves new state-of-the-art results on the CUB-200-2011 dataset.
Memorability determines what evanesces into emptiness, and what worms its way into the deepest furrows of our minds. It is the key to curating more meaningful media content as we wade through daily digital torrents. The Predicting Media Memorability task in MediaEval 2020 aims to address the question of media memorability by setting the task of automatically predicting video memorability. Our approach is a multimodal deep learning-based late fusion that combines visual, semantic, and auditory features. We used audio gestalt to estimate the influence of the audio modality on overall video memorability, and accordingly inform which combination of features would best predict a given video's memorability scores.
Memes on the Internet are often harmless and sometimes amusing. However, by using certain types of images, text, or combinations of both, the seemingly harmless meme becomes a multimodal type of hate speech -- a hateful meme. The Hateful Memes Challenge is a first-of-its-kind competition which focuses on detecting hate speech in multimodal memes and it proposes a new data set containing 10,000+ new examples of multimodal content. We utilize VisualBERT -- which meant to be the BERT of vision and language -- that was trained multimodally on images and captions and apply Ensemble Learning. Our approach achieves 0.811 AUROC with an accuracy of 0.765 on the challenge test set and placed third out of 3,173 participants in the Hateful Memes Challenge.
With expansion of the video advertising market, research to predict the effects of video advertising is getting more attention. Although effect prediction of image advertising has been explored a lot, prediction for video advertising is still challenging with seldom research. In this research, we propose a method for predicting the click through rate (CTR) of video advertisements and analyzing the factors that determine the CTR. In this paper, we demonstrate an optimized framework for accurately predicting the effects by taking advantage of the multimodal nature of online video advertisements including video, text, and metadata features. In particular, the two types of metadata, i.e., categorical and continuous, are properly separated and normalized. To avoid overfitting, which is crucial in our task because the training data are not very rich, additional regularization layers are inserted. Experimental results show that our approach can achieve a correlation coefficient as high as 0.695, which is a significant improvement from the baseline (0.487).
Current research in eSports lacks the tools for proper game practising and performance analytics. The majority of prior work relied only on in-game data for advising the players on how to perform better. However, in-game mechanics and trends are frequently changed by new patches limiting the lifespan of the models trained exclusively on the in-game logs. In this article, we propose the methods based on the sensor data analysis for predicting whether a player will win the future encounter. The sensor data were collected from 10 participants in 22 matches in League of Legends video game. We have trained machine learning models including Transformer and Gated Recurrent Unit to predict whether the player wins the encounter taking place after some fixed time in the future. For 10 seconds forecasting horizon Transformer neural network architecture achieves ROC AUC score 0.706. This model is further developed into the detector capable of predicting that a player will lose the encounter occurring in 10 seconds in 88.3% of cases with 73.5% accuracy. This might be used as a players' burnout or fatigue detector, advising players to retreat. We have also investigated which physiological features affect the chance to win or lose the next in-game encounter.
Reliable detection of the prodromal stages of Alzheimer's disease (AD) remains difficult even today because, unlike other neurocognitive impairments, there is no definitive diagnosis of AD in vivo. In this context, existing research has shown that patients often develop language impairment even in mild AD conditions. We propose a multimodal deep learning method that utilizes speech and the corresponding transcript simultaneously to detect AD. For audio signals, the proposed audio-based network, a convolutional neural network (CNN) based model, predicts the diagnosis for multiple speech segments, which are combined for the final prediction. Similarly, we use contextual embedding extracted from BERT concatenated with a CNN-generated embedding for classifying the transcript. The individual predictions of the two models are then combined to make the final classification. We also perform experiments to analyze the model performance when Automated Speech Recognition (ASR) system generated transcripts are used instead of manual transcription in the text-based model. The proposed method achieves 85.3% 10-fold cross-validation accuracy when trained and evaluated on the Dementiabank Pitt corpus.
Memes are used for spreading ideas through social networks. Although most memes are created for humor, some memes become hateful under the combination of pictures and text. Automatically detecting the hateful memes can help reduce their harmful social impact. Unlike the conventional multimodal tasks, where the visual and textual information is semantically aligned, the challenge of hateful memes detection lies in its unique multimodal information. The image and text in memes are weakly aligned or even irrelevant, which requires the model to understand the content and perform reasoning over multiple modalities. In this paper, we focus on multimodal hateful memes detection and propose a novel method that incorporates the image captioning process into the memes detection process. We conduct extensive experiments on multimodal meme datasets and illustrated the effectiveness of our approach. Our model achieves promising results on the Hateful Memes Detection Challenge.