Pedestrian Attribute Recognition with Part-based CNN and Combined Feature Representations

Chen, Yiqiang; Duffner, Stefan; Stoian, Andrei; Dufour, Jean-Yves; Başkurt, Atilla

doi:10.5220/0006622901140122

Cited by 15 publications

(19 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…age, gender).To go further, the framework allows an easy addition of new blocks, even for new features, in order to answer new use-cases, or to have better performance on already covered use-cases. A coming block, is the implementation of the work of our colleagues [3] to get 45 more features (e.g. Causal/Formal upper/lower clothes, carrying plastic bag, gender) of people.…”

Section: Resultsmentioning

confidence: 99%

See 1 more Smart Citation

People Management Framework Using a 2D Camera for Human-Robot Social Interactions

Saraydaryan

Leber

Jumel

2019

Lecture Notes in Computer Science

View full text Add to dashboard Cite

In order to perform tasks and offer socially acceptable humanrobot interactions, domestic robots need the ability to collect various information about people. In this paper, we propose a framework that allows the extraction of high-level person features from a 2D camera in addition to tracking people over time. The proposed people management framework aggregates body and person features including an original pose estimation using only a 2D camera. At this time, people pose and posture, clothing colors, face recognition are combined with tracking and re-identification abilities. This framework has been successfully used by the LyonTech team in the RoboCup@Home 2018 competition with a Pepper robot from SoftBank Robotics where its utility for domestic robot applications was demonstrated.

show abstract

Section: Resultsmentioning

confidence: 99%

“…In recent works, the Deep Learning approach applied on large dataset has allowed the extraction of numerous human features (e.g. clothes color, gender, hat, hair color) [2,3]. All these approaches give important clues and tools to characterize and track people and could be based on simple sensors such as a 2D camera.…”

Section: Introductionmentioning

confidence: 99%

People Management Framework Using a 2D Camera for Human-Robot Social Interactions

Saraydaryan

Leber

Jumel

2019

Lecture Notes in Computer Science

View full text Add to dashboard Cite

show abstract

“…In fact, many attributes just appear in small regions on human body, so using a unique global feature vector to learn all attributes is inefficient. Recognizing this drawback, [28,29] proposed part-based CNN models, in which a global feature map from a middle layer is horizontally split into 4 disjoint equal local feature maps, each one is then forwarded to some other convolutional layers followed by a last linear layer to predict probabilities of a group of attributes. The improvement in this method is that it use multiple local features to predict groups of suitable attributes.…”

Section: Attribute For Re-idmentioning

confidence: 99%

“…However, most attributes appear in local positions, so global features are insufficient to recognize them. Some works notice this drawback and improve by divide global features into local parts [28,29], but they still consider attributes as an auxiliary branch to enrich deep features.…”

Section: Introductionmentioning

confidence: 99%

Person Re-Identification System at Semantic Level based on Pedestrian Attributes Ontology

Ly¹,

Cao²,

Nguyen³

2020

IJACSA

View full text Add to dashboard Cite

Person Re-Identification (Re-ID) is a very important task in video surveillance systems such as tracking people, finding people in public places, or analysing customer behavior in supermarkets. Although there have been many works to solve this problem, there are still remaining challenges such as large-scale datasets, imbalanced data, viewpoint, finegrained data (attributes), the Local Features are not employed at semantic level in online stage of Re-ID task, furthermore, the imbalanced data problem of attributes are not taken into consideration. This paper has proposed a Unified Re-ID system consisted of three main modules such as Pedestrian Attribute Ontology (PAO), Local Multi-task DCNN (Local MDCNN), Imbalance Data Solver (IDS). The new main point of our Re-ID system is the power of mutual support of PAO, Local MDCNN and IDS to exploit the inner-group correlations of attributes and pre-filter the mismatch candidates from Gallery set based on semantic information as Fashion Attributes and Facial Attributes, to solve the imbalanced data of attributes without adjusting network architecture and data augmentation. We experimented on the well-known Market1501 dataset. The experimental results have shown the effectiveness of our Re-ID system and it could achieve the higher performance on Market1501 dataset in comparison to some state-of-the-art Re-ID methods.

show abstract

“…In order to learn the relationship between the mid-level features and the attributes, these detected layers are trained using image labels only by adopting a max-pooling based weakly-supervised object detection technique. Combining CNN extracted features with LOMO features, a part-based network is proposed by [16]. LOMO features, based on HSV histograms and Scale-Invariant Local Ternary Patterns, are texture and color descriptors shown to be illumination-invariant.…”

Section: Related Workmentioning

confidence: 99%

A multi-branch separable convolution neural network for pedestrian attribute recognition

Junejo

Ahmed

2020

Heliyon

View full text Add to dashboard Cite

Video surveillance applications have made great strides in making the world a safer place. Extracting visual attributes from a scene, such as the type of shoes, the type of clothing, carrying any object or not, or wearing any accessory etc., is a challenging problem and an efficient solution holds the key to a great number of applications. In this paper, we present a multi-branch convolutional neural network that uses depthwise separable convolution (DSC) layers to solve the pedestrian attribute recognition problem. Researchers have proposed various solutions over the years making use of convolutional neural networks (CNN), however, we introduce DSC layers to the CNN for the problem of pedestrian attribute recognition. In addition, we make a novel use of the different color spaces and create a 3-branch CNN, denoted as 3bCNN, that is efficient, especially with smaller datasets. We experiment on two benchmark datasets and show results with improvement over the state of the art.

show abstract

Pedestrian Attribute Recognition with Part-based CNN and Combined Feature Representations

Cited by 15 publications

References 0 publications

People Management Framework Using a 2D Camera for Human-Robot Social Interactions

People Management Framework Using a 2D Camera for Human-Robot Social Interactions

Person Re-Identification System at Semantic Level based on Pedestrian Attributes Ontology

A multi-branch separable convolution neural network for pedestrian attribute recognition

Contact Info

Product

Resources

About