Discrete online cross-modal hashing

Zhan, Yu-Wei; Wang, Yongxin; Sun, Youshun; Wu, Xiao-Ming; Luo, Xin; Xu, Xin-Shun

doi:10.1016/j.patcog.2021.108262

Cited by 49 publications

(5 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Within the rich landscape of multimodal image retrieval methods for retrieval, various strategies combining different modalities have been explored in the literature with more emphasis on hashing methods due to fast queries and less memory consumption. For instance, there are methods like Discrete Online Cross-modal Hashing (DOCH) [31], which creates high-quality hash codes for different data types by using both the likeness between data points and their detailed meanings. Fast Cross-Modal Hashing (FCMH) [26] adds an extra element to estimate the binary code, making it better by reducing mistakes.…”

Section: Multimodal Image Retrievalmentioning

confidence: 99%

Multimodal 3D Object Retrieval

Pegia,

Jónsson,

Moumtzidou

et al. 2024

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Three-dimensional (3D) retrieval of objects and models plays a crucial role in many application areas, such as industrial design, medical imaging, gaming and virtual and augmented reality. Such 3D retrieval involves storing and retrieving different representations of single objects, such as images, meshes or point clouds. Early approaches considered only one such representation modality, but recently the CMCL method has been proposed, which considers multimodal representations. Multimodal retrieval, meanwhile, has recently seen significant interest in the image retrieval domain. In this paper, we therefore explore the application of state-of-the-art multimodal image representations to 3D retrieval, in comparison to existing 3D approaches. In a detailed study over two benchmark 3D datasets, we show that the MuseHash approach from the image domain outperforms other approaches, improving recall over the CMCL approach by about 11% for unimodal retrieval and 9% for multimodal retrieval.

show abstract

Section: Multimodal Image Retrievalmentioning

confidence: 99%

Multimodal 3D Object Retrieval

Pegia,

Jónsson,

Moumtzidou

et al. 2024

Lecture Notes in Computer Science

View full text Add to dashboard Cite

show abstract

“…OLCH [31] introduces online semantic representation learning strategy to preserve the similarity between new data and old data in Hamming space. Besides, in order to avoid quantization error, DOCH [32] optimizes discretely binary constraints and yields uniform high-quality hash codes. OMGH [33] utilizes anchor-based manifold embedding to sparsely represent old data and adaptively guide hash learning.…”

Section: B Continuous Cross-modal Hashingmentioning

confidence: 99%

“…In order to objectively evaluate the performance of DLCH, we adopt two widely-used evaluations, including Mean Average Precision (MAP) and Precision-Recall Curves Hamming distance 2. Then, we compare our method with nine stateof-the-art hashing methods, including SSAH [34], DBRC [35], RDCMH [36], SADCH [37], MESDCH [38], OCMFH [27], OLSH [29], LEMON [30]and DOCH [32]. The first five methods are non-continuous deep cross-modal hashing methods and the rest are online cross-modal hashing methods.…”

Section: B Implementation Details and Evaluationsmentioning

confidence: 99%

Multi-Manifold Deep Discriminative Cross-Modal Hashing for Medical Image Retrieval

Zeng

Zheng

et al. 2022

IEEE Trans. on Image Process.

View full text Add to dashboard Cite

Hashing methods have made significant progress in cross-modal retrieval tasks with fast query speed and low storage cost. Among them, deep learning-based hashing achieves better performance on large-scale data due to its excellent extraction and representation ability for nonlinear heterogeneous features. However, there are still two main challenges in catastrophic forgetting when data with new categories arrive continuously, and time-consuming for non-continuous hashing retrieval to retrain for updating. To this end, we, in this paper, propose a novel deep lifelong cross-modal hashing to achieve lifelong hashing retrieval instead of re-training hash function repeatedly when new data arrive. Specifically, we design lifelong learning strategy to update hash functions by directly training the incremental data instead of retraining new hash functions using all the accumulated data, which significantly reduce training time. Then, we propose lifelong hashing loss to enable original hash codes participate in lifelong learning but remain invariant, and further preserve the similarity and dis-similarity among original and incremental hash codes to maintain performance. Additionally, considering distribution heterogeneity when new data arriving continuously, we introduce multi-label semantic similarity to supervise hash learning, and it has been proven that the similarity improves performance with detailed analysis. Experimental results on benchmark datasets show that the proposed methods achieves comparative performance comparing with recent state-of-the-art cross-modal hashing methods, and it yields substantial average increments over 20% in retrieval accuracy and almost reduces over 80% training time when new data arrives continuously.

show abstract

“…ing (FOMH) (Lu et al 2019a) learns the modal combination weights adaptively based on the online streaming multimodal data. Discrete Online Cross-modal Hashing (DOCH) (Zhan et al 2022) considers the fine-grained semantic information to learn the binary codes of the new data. In this paper, we focus on single-modal supervised online hashing tasks.…”

Section: Introductionmentioning

confidence: 99%

Fast Online Hashing with Multi-Label Projection

Jia

Cao

Liu

et al. 2023

AAAI

View full text Add to dashboard Cite

Hashing has been widely researched to solve the large-scale approximate nearest neighbor search problem owing to its time and storage superiority. In recent years, a number of online hashing methods have emerged, which can update the hash functions to adapt to the new stream data and realize dynamic retrieval. However, existing online hashing methods are required to update the whole database with the latest hash functions when a query arrives, which leads to low retrieval efficiency with the continuous increase of the stream data. On the other hand, these methods ignore the supervision relationship among the examples, especially in the multi-label case. In this paper, we propose a novel Fast Online Hashing (FOH) method which only updates the binary codes of a small part of the database. To be specific, we first build a query pool in which the nearest neighbors of each central point are recorded. When a new query arrives, only the binary codes of the corresponding potential neighbors are updated. In addition, we create a similarity matrix which takes the multi-label supervision information into account and bring in the multi-label projection loss to further preserve the similarity among the multi-label data. The experimental results on two common benchmarks show that the proposed FOH can achieve dramatic superiority on query time up to 6.28 seconds less than state-of-the-art baselines with competitive retrieval accuracy.

show abstract

Discrete online cross-modal hashing

Cited by 49 publications

References 18 publications

Multimodal 3D Object Retrieval

Multimodal 3D Object Retrieval

Multi-Manifold Deep Discriminative Cross-Modal Hashing for Medical Image Retrieval

Fast Online Hashing with Multi-Label Projection

Contact Info

Product

Resources

About