Cross-Modal Retrieval With CNN Visual Features: A New Baseline

By Yunchao Wei, Yao Zhao, Canyi Lu, Shikui Wei, Luoqi Liu, Zhenfeng Zhu, Shuicheng Yan. IEEE Transactions on Cybernetics, 2017

370 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/ce3c5b0691d9e05061adc0a6fed8acb2607b1423

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

1. Yi Yang
2. Liang Zheng
3. Xiaogang Wang
4. Dacheng Tao
5. Luc Van Gool
6. Yunchao Wei
7. Shaogang Gong
8. Wanli Ouyang

Citations in the corpus

• Dual-path Convolutional Image-Text Embeddings with Instance Loss (2017, background): “What motivates us is the effectiveness of class labels in earlier years of cross-media retrieval [58, 64, 65, 69, 71].”
• A Survey of Multi-View Representation Learning (2016, neutral): “[185] introduce a deep semantic matching method, in which two”
• Product1M: Towards Weakly Supervised Instance-Level Product Retrieval via Cross-Modal Pretraining (2021, background): “Cross-modal retrieval [43, 12, 48, 4, 44, 8] emerges as a promising avenue for efficient indexing and searching among large-scale data with different modalities, and is widely used in search engines [2, 14], E-commerce [18, 7], to name a few.”
• RGB-Infrared Cross-Modality Person Re-identification (2017, background): “one includes subspace learning methods [25, 30, 60] and deep learning frameworks [45, 6, 14, 12], while the latter one includes linear models [39, 36, 59, 51] and non-linear models [27, 50, 31].”
• A Decade Survey of Content Based Image Retrieval Using Deep Learning (2020, methodology): “Pairwise Labels [97], [98] Different Losses [65], [68], [71] CNN Features [112], [97], [134] 2017 Triplet + Siamese [72]”
• A Comprehensive Survey on Cross-modal Retrieval (2016, methodology): “Wei et al. [58] propose a deep semantic matching method to address the cross-modal retrieval problem for samples which are annotated with one or multiple labels.”
• Channel Augmented Joint Learning for Visible-Infrared Recognition (2021, background): “Learning modality-related metrics or dictionaries with hand-craft features [39, 18] is the popular approach.”
• Dual-Path Convolutional Image-Text Embedding (2017, methodology): “[38] finetunes the CNN by classifying broad categories on target datasets.”
• An End-to-End Steel Surface Defect Detection Approach via Fusing Multiple Hierarchical Features (2020, methodology): “Unsurprisingly, the performance of the methods based on convolutional features is strongly higher than the methods without CNN [37].”
• Deep Supervised Cross-Modal Retrieval (2019, background): “Inspired by the great success of deep neural networks in representation learning [14], a large number of deep learning-based approaches [2, 33, 19, 36, 21, 25, 7] have been proposed to learn a common presentation space for multimedia data.”
• An Overview of Cross-Media Retrieval: Concepts, Methodologies, Benchmarks, and Challenges (2017, background): “[40] propose deep-SM to exploit convolutional neural network (CNN) with deep semantic matching, which demonstrates the power of the CNN features in cross-media retrieval.”