11,112 Hits in 5.0 sec

Learning Intra-Batch Connections for Deep Metric Learning [article]

Jenny Seidenschwarz, Ismail Elezi, Laura Leal-Taixé
2021 arXiv   pre-print
Particularly, deep metric learning utilizes neural networks to learn such a mapping.  ...  The goal of metric learning is to learn a function that maps samples to a lower-dimensional space where similar samples lie closer than dissimilar ones.  ...  We thank Guillem Brasó for useful discussions.  ... 
arXiv:2102.07753v3 fatcat:ikholheukvbcxigdxv5cefjkpu

Deep Cross-modality Adaptation via Semantics Preserving Adversarial Learning for Sketch-based 3D Shape Retrieval [article]

Jiaxin Chen, Yi Fang
2018 arXiv   pre-print
Specifically, we first separately adopt two metric networks, following two deep convolutional neural networks (CNNs), to learn modality-specific discriminative features based on an importance-aware metric  ...  To address this problem, we propose a novel framework to learn a discriminative deep cross-modality adaptation model in this paper.  ...  Similar to most existing deep learning methods, we train our model by mini-batches.  ... 
arXiv:1807.01806v1 fatcat:xuww7nfkkncqdbszqcgf7sls3e

Deep ranking model by large adaptive margin learning for person re-identification

Jiayun Wang, Sanping Zhou, Jinjun Wang, Qiqi Hou
2018 Pattern Recognition  
In this paper, we present a novel deep ranking model with feature learning and fusion by learning a large adaptive margin between the intra-class distance and inter-class distance to solve the person re-identification  ...  Specifically, we organize the training images into a batch of pairwise samples.  ...  In this paper, we propose a novel deep ranking model with feature learning and fusion by learning a large adaptive margin between the intra-class samples and inter-class samples for person re-identification  ... 
doi:10.1016/j.patcog.2017.09.024 fatcat:zwoaj2col5aa5oxnfef6vfmpnu

Self-attention Learning for Person Re-identification

Minyue Jiang, Yuan Yuan, Qi Wang
2018 British Machine Vision Conference  
And we combine softmax loss with quadruplet loss, taking full advantages of labels and metric learning at the same time.  ...  To remedy this problem, we propose a novel yet simple self-attention learning method for person re-identification.  ...  Compared with end-to-end deep learning method, feature extraction and metric learning are independent. Deep learning methods have surpassed traditional hand-crafted designed methods.  ... 
dblp:conf/bmvc/Jiang0018 fatcat:2pjlp6fknvfglezsuncbwrxzhy

Visible Thermal Person Re-Identification via Dual-Constrained Top-Ranking

Mang Ye, Zheng Wang, Xiangyuan Lan, Pong C. Yuen
2018 Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence  
It is advantageous in two aspects: 1) end-to-end feature learning directly from the data without extra metric learning steps, 2) it simultaneously handles the cross-modality and intra-modality variations  ...  Cross-modality person re-identification between the thermal and visible domains is extremely important for night-time surveillance applications.  ...  presented a deep zero-padding network to learn the invariant feature representations. In contrast, we present an end-to-end dual-path learning framework for feature and metric learning.  ... 
doi:10.24963/ijcai.2018/152 dblp:conf/ijcai/YeWLY18 fatcat:hgcunneocnf3lfns3gnsfgpyge

Joint Learning of the Center Points and Deep Metrics for Land-Use Classification in Remote Sensing

Zhiqiang Gong, Ping Zhong, Weidong Hu, Yuming Hua
2019 Remote Sensing  
This work introduces structured metric learning for remote sensing scene representation, a special deep metric learning which can take full advantage of the training batch.  ...  To address this problem, deep metric learning, which incorporates the metric learning into the deep model, is used to maximize the inter-class variance and minimize the intra-class variance for better  ...  ., materials used for experiments). Conflicts of Interest: The authors declare no conflict of interest.  ... 
doi:10.3390/rs11010076 fatcat:ifjd7biirzbpnplr7gfczs243q

SoftTriple Loss: Deep Metric Learning Without Triplet Sampling [article]

Qi Qian, Lei Shang, Baigui Sun, Juhua Hu, Hao Li, Rong Jin
2020 arXiv   pre-print
Compared with conventional deep metric learning algorithms, optimizing SoftTriple loss can learn the embeddings without the sampling phase by mildly increasing the size of the last fully connected layer  ...  Due to the vast number of triplet constraints, a sampling strategy is essential for DML. With the tremendous success of deep learning in classifications, it has been applied for DML.  ...  Deep metric learning Deep metric learning aims to learn the embeddings directly from the raw materials (e.g., images) by deep neural networks [15, 21] .  ... 
arXiv:1909.05235v2 fatcat:wm7627l5q5eq5fphvrsw5oxan4

Class-Wise Centroid Distance Metric Learning for Acoustic Event Detection

Xugang Lu, Peng Shen, Sheng Li, Yu Tsao, Hisashi Kawai
2019 Interspeech 2019  
This pipeline is directly connected to a final goal for class discrimination without explicitly considering how the features should be distributed for inter-class and intra-class samples.  ...  In this paper, we explicitly add a distance metric constraint on feature extraction process with a goal to reduce intra-class sample distances and increase inter-class sample distances.  ...  Many models based on deep learning have been proposed for the AED [7, 8, 4] .  ... 
doi:10.21437/interspeech.2019-2271 dblp:conf/interspeech/LuS00K19 fatcat:5kpp6jvlx5hkviwbgthur54tgi

Learning With Batch-Wise Optimal Transport Loss for 3D Shape Recognition

Lin Xu, Han Sun, Yuai Liu
2019 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)  
Deep metric learning is essential for visual recognition.  ...  We use it to learn the distance metric and deep feature representation jointly for recognition.  ...  In this paper, we propose a novel batch-wise optimal transport loss objective for deep metric learning.  ... 
doi:10.1109/cvpr.2019.00345 dblp:conf/cvpr/XuSL19 fatcat:yms72ja3djfurctwhpexdg35aq

Deep Metric Learning with Density Adaptivity [article]

Yehao Li and Ting Yao and Yingwei Pan and Hongyang Chao and Tao Mei
2019 arXiv   pre-print
With the rise and success of Convolutional Neural Networks (CNN), deep metric learning (DML) involves training a network to learn a nonlinear transformation to the embedding space.  ...  The problem of distance metric learning is mostly considered from the perspective of learning an embedding space, where the distances between pairs of examples are in correspondence with a similarity metric  ...  [16] is one of the early works to capitalize on contrastive embedding for deep metric learning in face verification task.  ... 
arXiv:1909.03909v1 fatcat:b4dj5qns3jendd2n2n2cx5nquq

Learning with Batch-wise Optimal Transport Loss for 3D Shape Recognition [article]

Lin Xu, Han Sun, Yuai Liu
2019 arXiv   pre-print
Deep metric learning is essential for visual recognition.  ...  We use it to learn the distance metric and deep feature representation jointly for recognition.  ...  In this paper, we propose a novel batch-wise optimal transport loss objective for deep metric learning.  ... 
arXiv:1903.08923v1 fatcat:szahnuma45g2znkg4teqw5aavm

Dysarthric Speech Recognition Based on Deep Metric Learning

Yuki Takashima, Ryoichi Takashima, Tetsuya Takiguchi, Yasuo Ariki
2020 Interspeech 2020  
To alleviate this intra-class variation problem, we propose an ASR system based on deep metric learning.  ...  Experimental results show that our proposed approach using deep metric learning improves the word-recognition accuracy consistently.  ...  In this paper, we propose a deep metric learning-based ASR system that takes into consideration intra-class variation.  ... 
doi:10.21437/interspeech.2020-2267 dblp:conf/interspeech/TakashimaTTA20 fatcat:26roqfcqubfkdh2en2ajn4okbu

Deep adaptive feature embedding with local sample distributions for person re-identification

Lin Wu, Yang Wang, Junbin Gao, Xue Li
2018 Pattern Recognition  
To this end, a novel objective function is proposed to jointly optimize similarity metric learning, local positive mining and robust deep embedding.  ...  Our method is capable of learning a deep similarity metric adaptive to local sample structure by minimizing each sample's local distances while propagating through the relationship between samples to attain  ...  To jointly learn representations and similarity metric for pedestrian samples, deep embedding approaches are developed to allow the interaction between feature extraction and metric learning [19, 20,  ... 
doi:10.1016/j.patcog.2017.08.029 fatcat:xt65d6molnhbfpriyvjskyjeym

Unsupervised Domain Adaptation via Discriminative Manifold Embedding and Alignment [article]

You-Wei Luo, Chuan-Xian Ren, Pengfei Ge, Ke-Kun Huang, Yu-Feng Yu
2020 arXiv   pre-print
Second, the batch-wise training manner in deep learning limits the description of the global structure.  ...  Though deep learning and adversarial strategy make an important breakthrough in the adaptability of features, there are two issues to be further explored.  ...  Finally, the discriminative structure loss is noted by L DS = i (L i inter + L i intra ), Global Structure Learning For the batch scheme in deep models, it is hard to obtain the complete relation graph  ... 
arXiv:2002.08675v2 fatcat:wjnr7smuo5bvfjkrkg2ycqmedq

DarkRank: Accelerating Deep Metric Learning via Cross Sample Similarities Transfer [article]

Yuntao Chen, Naiyan Wang, Zhaoxiang Zhang
2017 arXiv   pre-print
This knowledge can be naturally derived from deep metric learning model. To transfer them, we bring the "learning to rank" technique into deep metric learning formulation.  ...  We test our proposed DarkRank method on various metric learning tasks including pedestrian re-identification, image retrieval and image clustering. The results are quite encouraging.  ...  The key point of metric learning is to separate inter-class embeddings and reduce the intra-class variance.  ... 
arXiv:1707.01220v2 fatcat:3bwxube6djd3nhxtg3nfexxpom
« Previous Showing results 1 — 15 out of 11,112 results