Deep Polarized Network for Supervised Learning of Accurate Binary Hashing Codes

Li, Fan; Ng, Kam Woh; Ju, Ce; Zhang, Tianyu; Chan, Chee Seng

doi:10.24963/ijcai.2020/115

Cited by 80 publications

(64 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…From our experiment, manipulating hyperparameters of the algorithm and batch size of input data are able to generate distorted images to fool the classification model which could lead other potential threats. Luckily, there are some existing works [16][17][18] which suggest different encryption methods in federated learning as a defence.…”

Section: Discussionmentioning

confidence: 99%

From Gradient Leakage To Adversarial Attacks In Federated Learning

Lim¹,

Chan

2021

2021 IEEE International Conference on Image Processing (ICIP)

Self Cite

View full text Add to dashboard Cite

Deep neural networks (DNN) are widely used in real-life applications despite the lack of understanding on this technology and its challenges. Data privacy is one of the bottlenecks that is yet to be overcome and more challenges in DNN arise when researchers start to pay more attention to DNN vulnerabilities. In this work, we aim to cast the doubts towards the reliability of the DNN with solid evidence particularly in Federated Learning environment by utilizing an existing privacy breaking algorithm which inverts gradients of models to reconstruct the input data. By performing the attack algorithm, we exemplify the data reconstructed from inverting gradients algorithm as a potential threat and further reveal the vulnerabilities of models in representation learning.

show abstract

Section: Discussionmentioning

confidence: 99%

From Gradient Leakage To Adversarial Attacks In Federated Learning

Lim¹,

Chan

2021

2021 IEEE International Conference on Image Processing (ICIP)

Self Cite

View full text Add to dashboard Cite

show abstract

“…[28] introduced a deep incremental hash framework that alleviates the burden of retraining the model when the gallery image set increases. [22] further introduces a deep polarized loss for the hamming code generation, obviating the need for an additional quantization loss. Despite the success, these methods are not tailored for the large-scale vehicle re-identification scenario and could only achieve sub-optimal performances, as is illustrated before.…”

Section: A Hash For Content Retrievalmentioning

confidence: 99%

“…For the supervised setting, we further incorporate DSH [48] which was among the first works attempting to learn discrete hamming hash codes with deep convolutional neural networks. In addition, we incorporate other recent methods for detailed comparisons and analyses including DHN [24], HashNet [11], DCH, [13] DPN [22]. Note that we adopt original source codes for SH, ITQ, DCH and DHN.…”

Section: Comparisons With State-of-the-artsmentioning

confidence: 99%

DVHN: A Deep Hashing Framework for Large-scale Vehicle Re-identification

Chen¹,

Zhang²,

Liu³

et al. 2021

Preprint

View full text Add to dashboard Cite

Vehicle re-identification, which seeks to match query vehicle images with tremendous gallery images , has been gathering proliferating momentum. Conventional methods generally perform re-identification tasks by representing vehicle images as real-valued feature vectors and then ranking the gallery images by computing the corresponding Euclidean distances. Despite achieving remarkable retrieval accuracy, these methods require tremendous memory and computation when the gallery set is large, making them inapplicable in real-world scenarios.In light of this limitation, in this paper, we make the very first attempt to investigate the integration of deep hash learning with vehicle re-identification. We propose a deep hash-based vehicle re-identification framework, dubbed DVHN, which substantially reduces memory usage and promotes retrieval efficiency while reserving nearest neighbor search accuracy. Concretely, DVHN directly learns discrete compact binary hash codes for each image by jointly optimizing the feature learning network and the hash code generating module. Specifically, we directly constrain the output from the convolutional neural network to be discrete binary codes and ensure the learned binary codes are optimal for classification. To optimize the deep discrete hashing framework, we further propose an alternating minimization method for learning binary similarity-preserved hashing codes. Extensive experiments on two widely-studied vehicle re-identification datasets-VehicleID and VeRi-have demonstrated the superiority of our method against the state-of-the-art deep hash methods. DVHN of 2048 bits can achieve 13.94% and 10.21% accuracy improvement in terms of mAP and Rank@1 for VehicleID (800) dataset. For VeRi, we achieve 35.45% and 32.72% performance gains for Rank@1 and mAP, respectively.

show abstract

“…Various learning objectives are developed to learn hash codes using a training dataset. The objective functions include 1) task learning objective which can be further categorized into pointwise [48,53,39,40,11,49], pairwise [4,24,22], triplet-wise [44,31], listwise [51] and unsupervised [26,13]; 2) quantization error minimization such as the loss designed to minimize the p-norm (usually p = 2) between continuous codes and hash codes; 3) code balancing [26,39]. We refer readers to learning to hashing surveys [43,42,10] for more detailed review.…”

Section: Related Workmentioning

confidence: 99%

“…Recently, deep hashing methods [47,22] have shown great improvements over conventional hashing methods [45,14,15,21,35,36,18]. Furthermore, deep hashing methods can be grouped by how the similarity of the learned hashing codes are measured, namely pointwise [48,53,39,11,49], pairwise [24,22,4,3], triplet-wise [44,31], or listwise [51]. Among them, pointwise methods have a O(N ) computational complexity, whilst the complexity of the others are of at least O(N 2 ) for N data points.…”

Section: Introductionmentioning

confidence: 99%

One Loss for All: Deep Hashing with a Single Cosine Similarity based Learning Objective

Tian¹,

Ng²,

Zhang³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

A deep hashing model typically has two main learning objectives: to make the learned binary hash codes discriminative and to minimize a quantization error. With further constraints such as bit balance and code orthogonality, it is not uncommon for existing models to employ a large number (>4) of losses. This leads to difficulties in model training and subsequently impedes their effectiveness. In this work, we propose a novel deep hashing model with only a single learning objective. Specifically, we show that maximizing the cosine similarity between the continuous codes and their corresponding binary orthogonal codes can ensure both hash code discriminativeness and quantization error minimization. Further, with this learning objective, code balancing can be achieved by simply using a Batch Normalization (BN) layer and multi-label classification is also straightforward with label smoothing. The result is an one-loss deep hashing model that removes all the hassles of tuning the weights of various losses. Importantly, extensive experiments show that our model is highly effective, outperforming the state-of-the-art multi-loss hashing models on three large-scale instance retrieval benchmarks, often by significant margins. Code is available at https://github.com/kamwoh/orthohash * equal contribution.

show abstract

Deep Polarized Network for Supervised Learning of Accurate Binary Hashing Codes

Cited by 80 publications

References 25 publications

From Gradient Leakage To Adversarial Attacks In Federated Learning

From Gradient Leakage To Adversarial Attacks In Federated Learning

DVHN: A Deep Hashing Framework for Large-scale Vehicle Re-identification

One Loss for All: Deep Hashing with a Single Cosine Similarity based Learning Objective

Contact Info

Product

Resources

About