Collaborative Deep Networks for Pedestrian Detection

Song, Hongmeng; Wang, Wenmin; Wang, Jinzhuo; Wang, Ronggang

doi:10.1109/bigmm.2017.17

Cited by 8 publications

(3 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Tian et al [23] jointly optimized pedestrian detection with semantic tasks, including pedestrian attributes and scene attributes. Song et al [5] combined multiple deep networks with one fully-connected layer to improve the detection accuracy. In [24], CNN features extracted by a region proposal network (RPN) [11] are fed into the random forest for pedestrian detection.…”

Section: Pedestrian Detectionmentioning

confidence: 99%

“…For example, in pedestrian detection and person re-identification problems, camera views, poses, occlusions, illuminations, backgrounds and resolutions may easily cause intra-class dissimilarity and inter-class similarity. Therefore, both detection and retrieval are challenging problems in computer vision and have attracted lots of attention in recent years [3], [4], [5], [6], [7], [8].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Tasks Integrated Networks: Joint Detection and Retrieval for Image Search

Zhang

Yang

et al. 2022

IEEE Trans. Pattern Anal. Mach. Intell.

View full text Add to dashboard Cite

The traditional object (person) retrieval (re-identification) task aims to learn a discriminative feature representation with intra-similarity and inter-dissimilarity, which supposes that the objects in an image are manually or automatically pre-cropped exactly. However, in many real-world searching scenarios (e.g., video surveillance), the objects (e.g., persons, vehicles, etc.) are seldom accurately detected or annotated. Therefore, object-level retrieval becomes intractable without bounding-box annotation, which leads to a new but challenging topic, i.e. image-level search with multi-task integration of joint detection and retrieval. In this paper, to address the image search issue, we first introduce an end-to-end Integrated Net (I-Net), which has three merits: 1) A Siamese architecture and an on-line pairing strategy for similar and dissimilar objects in the given images are designed. Benefited by the Siamese structure, I-Net learns the shared feature representation, because, on which, both object detection and classification tasks are handled. 2) A novel on-line pairing (OLP) loss is introduced with a dynamic feature dictionary, which alleviates the multi-task training stagnation problem, by automatically generating a number of negative pairs to restrict the positives. 3) A hard example priority (HEP) based softmax loss is proposed to improve the robustness of classification task by selecting hard categories. The shared feature representation of I-Net may restrict the task-specific flexibility and learning capability between detection and retrieval tasks. Therefore, with the philosophy of divide and conquer, we further propose an improved I-Net, called DC-I-Net, which makes two new contributions: 1) two modules are tailored to handle different tasks separately in the integrated framework, such that the task specification is guaranteed. 2) A class-center guided HEP loss (C 2 HEP) by exploiting the stored class centers is proposed, such that the intra-similarity and inter-dissimilarity can be captured for ultimate retrieval. Extensive experiments on famous image-level search oriented benchmark datasets, such as CUHK-SYSU dataset and PRW dataset for person search and the large-scale WebTattoo dataset for tattoo search, demonstrate that the proposed DC-I-Net outperforms the state-of-the-art tasks-integrated and tasks-separated image search models.

show abstract

Section: Pedestrian Detectionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Tasks Integrated Networks: Joint Detection and Retrieval for Image Search

Zhang

Yang

et al. 2022

IEEE Trans. Pattern Anal. Mach. Intell.

View full text Add to dashboard Cite

show abstract

“…In recent years, CNN has achieved remarkable achievements in face recognition [ 17 ], handwritten digit recognition [ 18 ], pedestrian detection [ 19 ], and other fields, bringing new opportunities for the development of rice DOM detection technology. In terms of DOM detection of rice, Qi et al [ 20 ] combined the hypercolumn technology, max-relevance and min-redundancy feature selection algorithm, extreme learning machine technique, and improved VGG16 to identify rice DOM with an overall accuracy of 97.32%.…”

Section: Introductionmentioning

confidence: 99%

Evaluation of Rice Degree of Milling Based on Bayesian Optimization and Multi-Scale Residual Model

Chen

Wang

2022

Foods

View full text Add to dashboard Cite

Traditional machine learning-based methods for the detection of rice degree of milling (DOM) that are not comprehensive in feature extraction and have low recognition rates fail to meet the demand for fast, non-destructive, and accurate detection. This paper presents a digital image processing technology combined with deep learning to implement the classification of DOM of rice. An improved multi-scale information fusion model of the InceptionResNet–Bayesian optimization algorithm (IRBOA) was constructed based on the Inception-v3 structure and residual network (ResNet) model. It enables to automatically extract more comprehensive features of rice and determine the DOM of rice. Additionally, the important hyperparameters in the model were tuned by the BOA to optimize the recognition rate of rice DOM. The results show the hyperparameters optimized using the BOA are those that would not be chosen in manual tuning. The classification precision of the IRBOA model reached 99.22%, 94.92%, and 96.55% for well-milled, reasonably well-milled, and substandard rice, respectively, with an average accuracy of no less than 96.90%. This model improved 7.41% over the traditional machine learning model and at least 1.35% over the fashionable CNN model with strong generalization performance. This method effectively completes rapid, non-destructive, and accurate intelligent detection of rice DOM, which can supply a reliable and accurate technical mean for rice processing enterprises to guide the rice processing process.

show abstract

End-to-End Detection and Re-identification Integrated Net for Person Search

Zhang

2019

Computer Vision – ACCV 2018

View full text Add to dashboard Cite

This paper proposes a pedestrian detection and reidentification (re-id) integration net (I-Net) in an end-to-end learning framework. The I-Net is used in real-world video surveillance scenarios, where the target person needs to be searched in the whole scene videos, while the annotations of pedestrian bounding boxes are unavailable. By comparing to the successful CVPR'17 work [Xiao et al., 2017] for joint detection and re-id, we have three distinct contributions. First, we introduce a Siamese architecture of I-Net instead of 1 stream, such that a verification task can be implemented. Second, we propose a novel on-line pairing loss (OLP) and hard example priority softmax loss (HEP), such that only the hard negatives are posed much attention in loss computation. Third, an on-line dictionary for negative samples storage is designed in I-Net without recording the positive samples. We show our result on person search datasets, the gap between detection and re-identification is narrowed. The superior performance can be achieved.

show abstract

Collaborative Deep Networks for Pedestrian Detection

Cited by 8 publications

References 21 publications

Tasks Integrated Networks: Joint Detection and Retrieval for Image Search

Tasks Integrated Networks: Joint Detection and Retrieval for Image Search

Evaluation of Rice Degree of Milling Based on Bayesian Optimization and Multi-Scale Residual Model

End-to-End Detection and Re-identification Integrated Net for Person Search

Contact Info

Product

Resources

About