Multimodal Deep Learning
66 papers with code • 1 benchmarks • 17 datasets
Multimodal deep learning is a type of deep learning that combines information from multiple modalities, such as text, image, audio, and video, to make more accurate and comprehensive predictions. It involves training deep neural networks on data that includes multiple types of information and using the network to make predictions based on this combined data.
One of the key challenges in multimodal deep learning is how to effectively combine information from multiple modalities. This can be done using a variety of techniques, such as fusing the features extracted from each modality, or using attention mechanisms to weight the contribution of each modality based on its importance for the task at hand.
Multimodal deep learning has many applications, including image captioning, speech recognition, natural language processing, and autonomous vehicles. By combining information from multiple modalities, multimodal deep learning can improve the accuracy and robustness of models, enabling them to perform better in real-world scenarios where multiple types of information are present.
Datasets
Most implemented papers
DeepSeek-VL: Towards Real-World Vision-Language Understanding
The DeepSeek-VL family (both 1. 3B and 7B models) showcases superior user experiences as a vision-language chatbot in real-world applications, achieving state-of-the-art or competitive performance across a wide range of visual-language benchmarks at the same model size while maintaining robust performance on language-centric benchmarks.
XFlow: Cross-modal Deep Neural Networks for Audiovisual Classification
Our work improves on existing multimodal deep learning algorithms in two essential ways: (1) it presents a novel method for performing cross-modality (before features are learned from individual modalities) and (2) extends the previously proposed cross-connections which only transfer information between streams that process compatible data.
Learn to Combine Modalities in Multimodal Deep Learning
Combining complementary information from multiple modalities is intuitively appealing for improving the performance of learning-based approaches.
Multimodal Age and Gender Classification Using Ear and Profile Face Images
Experimental results indicated that profile face images contain a rich source of information for age and gender classification.
Audio-Conditioned U-Net for Position Estimation in Full Sheet Images
The goal of score following is to track a musical performance, usually in the form of audio, in a corresponding score representation.
Predicting the Leading Political Ideology of YouTube Channels Using Acoustic, Textual, and Metadata Information
Our analysis shows that the use of acoustic signal helped to improve bias detection by more than 6% absolute over using text and metadata only.
Analysis of Social Media Data using Multimodal Deep Learning for Disaster Response
Multimedia content in social media platforms provides significant information during disaster events.
HYDRA: A multimodal deep learning framework for malware classification
While traditional machine learning methods for malware detection largely depend on hand-designed features, which are based on experts’ knowledge of the domain, end-to-end learning approaches take the raw executable as input, and try to learn a set of descriptive features from it.
Image Search With Text Feedback by Visiolinguistic Attention Learning
In this work, we tackle this task by a novel Visiolinguistic Attention Learning (VAL) framework.
More Diverse Means Better: Multimodal Deep Learning Meets Remote Sensing Imagery Classification
In particular, we also investigate a special case of multi-modality learning (MML) -- cross-modality learning (CML) that exists widely in RS image classification applications.