Learning Consistent Feature Representation for Cross-Modal Multimedia Retrieval

By Cuicui Kang, Shiming Xiang, Shengcai Liao, Changsheng Xu, Chunhong Pan. 2015

213 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition, Information Retrieval & Search, Multimodal Learning

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

In Information Retrieval & Search

In Multimodal Learning

Citations in the corpus

Rescaling Egocentric Vision: Collection, Pipeline and Challenges for EPIC-KITCHENS-100 (2020, methodology): “mAP has been used in other retrieval works [98,100,101,102], yet it only considers relevance between items to be binary.”
Robust Unsupervised Cross-modal Hashing for Multimedia Retrieval (2020, background): “To model the relationship among heterogeneous multimedia data, the most popular strategy is to learn a common latent subspace where the similarity between learned features can be directly compared [6, 7, 25, 38, 40].”
Learning Discriminative Binary Codes for Large-scale Cross-modal Retrieval (2017, methodology): “[4] utilized the class labels with a dragging matrix to learn consistent feature representations for two modalities to facilitate the cross-modal matching, and Deng et al.”
An Overview of Cross-Media Retrieval: Concepts, Methodologies, Benchmarks, and Challenges (2017, methodology): “as [85] where the 399-dimensional word frequency features are used for texts, and the 512-dimensional GIST features are used for images.”
Cross-modal deep metric learning with multi-task regularization (2017, methodology): “Kang et al. propose Local Group based Consistent Feature Learning (LGCFL) [6], which adopts a local group-based priori to learn basis matrices of different modalities.”
Learning Cross-Modal Common Representations by Private–Shared Subspaces Separation (2020, methodology): “…approach is compared with 15 state-of-the-art methods, including ﬁve traditional methods, namely, CCA [8], CFA [25], KCCA [22], JRL [28], and LGCFL [62]; and ten DNN-based methods, namely, Corr-AE [17], DCCA [31], CMDH [14], DeepSM [63], ACMR [7], CCL [15], CBT [59], MCSM [11], CM-GAN [16], and…”
Know More Say Less: Image Captioning Based on Scene Graphs (2019, background): “plications such as semantic image search [8]–[10], vision-and-”
CKD: Cross-Task Knowledge Distillation for Text-to-Image Synthesis (2020, background): “An effective solution is cross-modal retrieval [1]–[5] which can retrieve the instances of both image”