Looking beyond appearances: Synthetic training data for deep CNNs in re-identification

Barbosa, Igor Barros; Cristani, Marco; Caputo, Barbara; Rognhaugen, Aleksander; Theoharis, Theoharis

doi:10.1016/j.cviu.2017.12.002

Cited by 171 publications

(119 citation statements)

References 47 publications

Supporting

Mentioning

118

Contrasting

Order By: Relevance

“…Naturally, the proposed multi-level representation is much richer and meaningful than the descriptor one, so the explanation. As well, when compared to the deep learning methods [7,35,36,37,50,55,57,58,59] on Market-1501 and CUHK03 datasets, the proposed method achieves better re-ID rates without needing neither data augmentation nor drop out or GPU computing.…”

Section: Empirical Analysis Of the Proposed Methods 1) Impact Of Bimentioning

confidence: 93%

“…Nevertheless, the study of the CNN model has started only recently in the person re-ID task [7,35,50,55,57,58,59] and that is due to the small scale of the existent re-ID datasets. Verification models treat person re-ID as a two-class recognition task, by taking a pair of images as input and determining whether they belong to the same person or not.…”

Section: B Deep Cnn Learningmentioning

confidence: 99%

See 1 more Smart Citation

Person re-ID while Crossing Different Cameras: Combination of Salient-Gaussian Weighted BossaNova and Fisher Vector Encodings

Mejdoub¹,

Ksibi²,

Amar³

et al. 2017

ijacsa

View full text Add to dashboard Cite

Abstract-Person re-identification (re-ID) is a challenging task in the camera surveillance field, since it addresses the problem of re-identifying people across multiple non-overlapping cameras. Most of existing approaches have been concentrated on: 1) achieving a robust and effective feature representation; and 2) enforcing discriminative metric learning to predict if two images represent the same identity. In this context, we present a new approach for person re-ID built upon multi-level descriptors. This is achieved by combining three complementary representations: salient-Gaussian Fisher Vector (SGFV) encoding method, salient-Gaussian BossaNova (SGBN) histogram encoding method and deep Convolutional Neural Network (CNN) features. The two first methods adapt the histogram encoding framework to the person re-ID task. This is achieved by integrating the pedestrian saliency map and the spatial location information, in the histogram encoding process. On one hand, human saliency is reliable and distinctive in the person re-ID task, since it can model the uniqueness of the identity. On the other hand, localizing a person in the image can effectively discard noisy background information. Finally, one of the most advanced metric learning in person re-ID: the Cross-view Quadratic Discriminant Analysis (XQDA) is applied on the top of the resulting description. The proposed method yields promising person re-ID results on two challenging image-based person re-ID benchmarks: CUHK03 and Market-1501.

show abstract

Section: Empirical Analysis Of the Proposed Methods 1) Impact Of Bimentioning

confidence: 93%

Section: B Deep Cnn Learningmentioning

confidence: 99%

Person re-ID while Crossing Different Cameras: Combination of Salient-Gaussian Weighted BossaNova and Fisher Vector Encodings

Mejdoub¹,

Ksibi²,

Amar³

et al. 2017

ijacsa

View full text Add to dashboard Cite

show abstract

“…For human bodies, some works [38,39] render images with the non-learned artist-defined MakeHuman model [40] for 3D pose estimation or person re-identification, correspondingly. However, statistical parametric models learned from 3D scans of a big human population, like SMPL [41], capture the real distribution of human body shape.…”

Section: Related Workmentioning

confidence: 99%

Learning Multi-human Optical Flow

et al. 2020

View full text Add to dashboard Cite

The optical flow of humans is well known to be useful for the analysis of human action. Recent optical flow methods focus on training deep networks to approach the problem. However, the training data used by them does not cover the domain of human motion. Therefore, we develop a dataset of multi-human optical flow and train optical flow networks on this dataset. We use a 3D model of the human body and motion capture data to synthesize realistic flow fields in both single-and multi-person images. We then train optical flow networks to estimate human flow fields from pairs of images. We demonstrate that our trained networks are more accurate than a wide range of top methods on heldout test data and that they can generalize well to real image sequences. The code, trained models and the dataset are available for research.

show abstract

“…1(d), photo-realistic images are included in the ElBa dataset. Training a model with synthetic data is a common practice [27,3] and annotations for texels are easily made available as an output of the image generation process. Layout attributes and individual ones (addressing the single texel can be varied in our proposed parametric synthesis model.…”

Section: Elba: Element-based Texture Datasetmentioning

confidence: 99%

Texture Retrieval in the Wild Through Detection-Based Attributes

Joppi

Godi

Giachetti

et al. 2019

Lecture Notes in Computer Science

Self Cite

View full text Add to dashboard Cite

Capturing the essence of a textile image in a robust way is important to retrieve it in a large repository, especially if it has been acquired in the wild (by taking a photo of the textile of interest). In this paper we show that a texel-based representation fits well with this task. In particular, we refer to Texel-Att, a recent texel-based descriptor which has shown to capture fine grained variations of a texture, for retrieval purposes. After a brief explanation of Texel-Att, we will show in our experiments that this descriptor is robust to distortions resulting from acquisitions in the wild by setting up an experiment in which textures from the ElBa (an Element-Based texture dataset) are artificially distorted and then used to retrieve the original image. We compare our approach with existing descriptors using a simple ranking framework based on distance functions. Results show that even under extreme conditions (such as a down-sampling with a factor of 10), we perform better than alternative approaches.

show abstract

Looking beyond appearances: Synthetic training data for deep CNNs in re-identification

Cited by 171 publications

References 47 publications

Person re-ID while Crossing Different Cameras: Combination of Salient-Gaussian Weighted BossaNova and Fisher Vector Encodings

Person re-ID while Crossing Different Cameras: Combination of Salient-Gaussian Weighted BossaNova and Fisher Vector Encodings

Learning Multi-human Optical Flow

Texture Retrieval in the Wild Through Detection-Based Attributes

Contact Info

Product

Resources

About