2019
DOI: 10.1016/j.patcog.2019.05.028
|View full text |Cite
|
Sign up to set email alerts
|

AlignedReID++: Dynamically matching local information for person re-identification

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
76
0

Year Published

2019
2019
2022
2022

Publication Types

Select...
4
4
1

Relationship

0
9

Authors

Journals

citations
Cited by 200 publications
(76 citation statements)
references
References 9 publications
0
76
0
Order By: Relevance
“…The contribution of the existing methods mainly vary depending on the problematic part they propose to solve. For example, [18] focus on addressing misalignment in person bounding boxes. In [2] multi-scale feature learning is adapted to include information about the small details at representation level.…”
Section: Deep Learning For Person Re-identificationmentioning
confidence: 99%
“…The contribution of the existing methods mainly vary depending on the problematic part they propose to solve. For example, [18] focus on addressing misalignment in person bounding boxes. In [2] multi-scale feature learning is adapted to include information about the small details at representation level.…”
Section: Deep Learning For Person Re-identificationmentioning
confidence: 99%
“…The performance of our proposed person Re-ID method is compared with the state-of-the-art methods on both Market-1501 and DukeMTMC-reID datasets. The employed comparison methods include AlignedReID [45], IDE (ID-discriminative embedding) [39], SVDNet (singular vector decomposition net) [46], TriNet (triplet net) [30], Pyramid [47], AWTL (adaptive weighted triplet loss) [48], ABD-Net (Attentive but Diverse Net) [49], DSA-reID (Densely Semantically Aligned reID) [50], the baseline method in [51], and the baseline method together with triplet loss [51].…”
Section: A Person Re-id Performancementioning
confidence: 99%
“…For instance, GoogleNet can provide the base knowledge for a multilevel triplet deep learning model to compute multilevel features efficiently for person re-identification tasks [13]. Similarly, models which use global features can benefit from the transferred knowledge for mapping local features to person reidentification tasks [14]. While fine-tuning works well on new tasks, it suffers forgetting problem on old tasks due to the absence of updating the weights for the original datasets.…”
Section: Related Workmentioning
confidence: 99%