Self-Supervised Feature Augmentation for Large Image Object Detection

Cite This

Files in this item

Files Size Format View

There are no files associated with this item.

PAN, Xingjia, Fan TANG, Weiming DONG, Yang GU, Zhichao SONG, Yiping MENG, Pengfei XU, Oliver DEUSSEN, Changsheng XU, 2020. Self-Supervised Feature Augmentation for Large Image Object Detection. In: IEEE Transactions on Image Processing. IEEE. 29, pp. 6745-6758. ISSN 1057-7149. eISSN 1941-0042. Available under: doi: 10.1109/TIP.2020.2993403

@article{Pan2020SelfS-49793, title={Self-Supervised Feature Augmentation for Large Image Object Detection}, year={2020}, doi={10.1109/TIP.2020.2993403}, volume={29}, issn={1057-7149}, journal={IEEE Transactions on Image Processing}, pages={6745--6758}, author={Pan, Xingjia and Tang, Fan and Dong, Weiming and Gu, Yang and Song, Zhichao and Meng, Yiping and Xu, Pengfei and Deussen, Oliver and Xu, Changsheng} }

<rdf:RDF xmlns:dcterms="" xmlns:dc="" xmlns:rdf="" xmlns:bibo="" xmlns:dspace="" xmlns:foaf="" xmlns:void="" xmlns:xsd="" > <rdf:Description rdf:about=""> <dc:creator>Dong, Weiming</dc:creator> <void:sparqlEndpoint rdf:resource="http://localhost/fuseki/dspace/sparql"/> <dc:language>eng</dc:language> <dc:creator>Xu, Changsheng</dc:creator> <dc:creator>Song, Zhichao</dc:creator> <dc:rights>terms-of-use</dc:rights> <dc:date rdf:datatype="">2020-06-05T05:53:55Z</dc:date> <dc:contributor>Pan, Xingjia</dc:contributor> <dcterms:abstract xml:lang="eng">Input scale plays an important role in modern detection frameworks, and an optimal training scale for images exists empirically. However, the optimal one usually cannot be reached in extremely large images under the memory constraint. In this study, we explore the scale effect inside the object detection pipeline and find that feature upsampling with the introduction of high-resolution information benefits the detection. Compared with direct input upscaling, feature upsampling trades a small performance loss for a large amount of memory savings. From these observations, we propose a self-supervised feature augmentation network, which takes downsampled images as inputs and aims to generate comparable features with the ones when feeding upscaled images to networks. We present a guided feature upsampling module, which takes downsampled images as inputs, to learn upscaled feature representations with the supervision of real large features acquired from upscaled images. In a self-supervised learning manner, we can introduce detailed information of images to the network. For an efficient feature upsampling, we design a residualized sub-pixel convolution block based on a sub-pixel convolution layer, which involves considerable information in upsampling process. Experiments on Mapillary Vistas Dataset (MVD), Cityscapes, and COCO are conducted to demonstrate the effectiveness of our method. On the MVD and Cityscapes detection benchmarks, in which the images are extremely large, our method surpasses current approaches. On COCO, the proposed method obtains comparable results to existing methods but with higher efficiency.</dcterms:abstract> <dc:creator>Tang, Fan</dc:creator> <dc:contributor>Meng, Yiping</dc:contributor> <dcterms:isPartOf rdf:resource=""/> <dc:creator>Meng, Yiping</dc:creator> <foaf:homepage rdf:resource="http://localhost:8080/jspui"/> <dc:creator>Xu, Pengfei</dc:creator> <dcterms:issued>2020</dcterms:issued> <dc:contributor>Xu, Pengfei</dc:contributor> <dc:contributor>Tang, Fan</dc:contributor> <dc:creator>Gu, Yang</dc:creator> <dcterms:available rdf:datatype="">2020-06-05T05:53:55Z</dcterms:available> <dspace:isPartOfCollection rdf:resource=""/> <dc:creator>Pan, Xingjia</dc:creator> <bibo:uri rdf:resource=""/> <dcterms:title>Self-Supervised Feature Augmentation for Large Image Object Detection</dcterms:title> <dc:creator>Deussen, Oliver</dc:creator> <dcterms:rights rdf:resource=""/> <dc:contributor>Xu, Changsheng</dc:contributor> <dc:contributor>Deussen, Oliver</dc:contributor> <dc:contributor>Gu, Yang</dc:contributor> <dc:contributor>Song, Zhichao</dc:contributor> <dc:contributor>Dong, Weiming</dc:contributor> </rdf:Description> </rdf:RDF>

This item appears in the following Collection(s)

Search KOPS


My Account