Content-Based Video-Music Retrieval Using Soft Intra-Modal Structure Constraint

Abstract

Up to now, only limited research has been conducted on cross-modal retrieval of suitable music for a specified video or vice versa. Moreover, much of the existing research relies on metadata such as keywords, tags, or associated description that must be individually produced and attached posterior. This paper introduces a new content-based, cross-modal retrieval method for video and music that is implemented through deep neural networks. We train the network via inter-modal ranking loss such that videos and music with similar semantics end up close together in the embedding space. However, if only the inter-modal ranking constraint is used for embedding, modality-specific characteristics can be lost. To address this problem, we propose a novel soft intra-modal structure loss that leverages the relative distance relationship between intra-modal samples before embedding. We also introduce reasonable quantitative and qualitative experimental protocols to solve the lack of standard protocols for less-mature video-music related tasks. Finally, we construct a large-scale 200K video-music pair benchmark. All the datasets and source code can be found in our online repository (https://github.com/csehong/VM-NET).

arXiv comment13 pages, 9 figures, 4 tables, supplementary material link >> https://youtu.be/ZyINqDMo3Fg

References in corpus (10)

• Adam: A Method for Stochastic Optimization
• Unifying Visual-Semantic Embeddings with Multimodal Neural Language Models
• YouTube-8M: A Large-Scale Video Classification Benchmark
• Convolutional Recurrent Neural Networks for Music Classification
• Deep Pyramidal Residual Networks
• Image-Text Multi-Modal Representation Learning by Adversarial Backpropagation
• Learning Two-Branch Neural Networks for Image-Text Matching Tasks
• SSPP-DAN: Deep Domain Adaptation Network for Face Recognition with Single Sample Per Person
• Beyond Deep Residual Learning for Image Restoration: Persistent Homology-Guided Manifold Simplification
• Recognizing Dynamic Scenes with Deep Dual Descriptor based on Key Frames and Key Segments

Cited by in corpus (2)

• Multi-Modal Music Information Retrieval: Augmenting Audio-Analysis with Visual Computing for Improved Music Video Analysis
• TriBERT: Full-body Human-centric Audio-visual Representation Learning for Visual Sound Separation