217 research outputs found
Galaxy classification with deep convolutional neural networks
No full textGalaxy classification, using digital images captured from sky surveys to determine the galaxy morphological classes, is of great interest to astronomy researchers. Conventional methods rely heavily on a few handcrafted morphological features while popular feature extraction methods that developed for natural images are not suitable for galaxy images. Deep convolutional neural networks (CNNs) are able to learn powerful features from images by hierarchical convolutional and pooling operations. This work applies state-of-the-art deep CNN technologies to galaxy classification for both a regression task and multi-class classification tasks. We also implement and compare the performance with several different conventional machine learning algorithms for a classification sub-task. Our experiments show that convolutional neural networks are able to learn representative features automatically and achieve high performance, surpassing both human recognition and other machine learning methods.Submission published under a 24 month embargo labeled 'Closed Access', the embargo will last until 2018-05-01The student, Honghui Shi, accepted the attached license on 2016-04-20 at 15:32.The student, Honghui Shi, submitted this Thesis for approval on 2016-04-20 at 15:46.This Thesis was approved for publication on 2016-04-26 at 09:28.DSpace SAF Submission Ingestion Package generated from Vireo submission #9386 on 2016-07-07 at 14:17:35Made available in DSpace on 2016-07-07T21:17:52Z (GMT). No. of bitstreams: 2
SHI-THESIS-2016.pdf: 4254004 bytes, checksum: 09e485fb4f8169bab7e4c553a1af319b (MD5)
LICENSE.txt: 4208 bytes, checksum: a3412add2c68bc23ddeee8497bb1ef01 (MD5)
Previous issue date: 2016-04-26Embargo set by: Seth Robbins for item 93294
Lift date: 2018-07-07T21:18:16Z
Reason: Author requested closed access (OA after 2yrs) in Vireo ETD systemLimited Restriction Lifted for Item 93294 on 2018-07-08T09:15:20Z
Deep learning in sequential data analysis
No full textDeep learning has achieved great success in recent years in computer vision and its related areas. For core computer vision tasks such as image classification, image semantic segmentation, image super-resolution, and object detection from images, deep learning based methods outperform various traditional methods in terms of both accuracy and speed. While a myriad of deep learning based computer vision research projects are continuously pushing forward the frontier of computer vision further by improving the performance for image-level tasks, many recent investigations have begun to look into deep learning based methods for sequential data such as videos and medical image sequences.
With the extra information from its additional sequential dimension, sequential data naturally raises an important and challenging question: How can we effectively and efficiently integrate such sequential information into existing successful and sophisticated image-based deep learning frameworks without building from scratch?
In this dissertation we develop techniques and methods that enable us to incorporate sequential information into existing image-based deep learning frameworks for different computer vision tasks. More specifically, we propose advanced methods that successfully utilize both image-based deep learning models and sequential information for the super-resolution task using multi-slice computed tomography image sequences, and for the object detection and tracking task using multi-frame videos. We demonstrate how we integrate sequential information into modern image-based deep learning systems for these different tasks under different integration paradigms. Our experiments show that our proposed methods have significantly improved the performances compared with naive image-based methods, and achieved the new state-of-the-art for such sequential vision tasks.Submission published under a 24 month embargo labeled 'Closed Access', the embargo will last until 2019-12-01The student, Honghui Shi, accepted the attached license on 2017-12-05 at 09:43.The student, Honghui Shi, submitted this Dissertation for approval on 2017-12-05 at 09:48.This Dissertation was approved for publication on 2017-12-05 at 11:50.DSpace SAF Submission Ingestion Package generated from Vireo submission #11851 on 2018-03-13 at 10:37:39Made available in DSpace on 2018-03-13T17:35:49Z (GMT). No. of bitstreams: 2
SHI-DISSERTATION-2017.pdf: 8966931 bytes, checksum: f7ebaa0ce2ebb11664a08f629fe9bd99 (MD5)
LICENSE.txt: 4208 bytes, checksum: d265a58770ecddde50f98f0b083e23db (MD5)
Previous issue date: 2017-12-05Embargo set by: Seth Robbins for item 105482
Lift date: 2020-03-13T17:36:05Z
Reason: Author requested closed access (OA after 2yrs) in Vireo ETD systemLimited Restriction Lifted for Item 105482 on 2020-03-14T09:15:28Z
Correction: Broadband ultrafast photovoltaic detectors based on large-scale topological insulator Sb<sub>2</sub>Te<sub>3</sub>/STO heterostructures
No full textCorrection for ‘Broadband ultrafast photovoltaic detectors based on large-scale topological insulator Sb2Te3/STO heterostructures’ by Honghui Sun, et al., Nanoscale, 2017, 9, 9325–9332.</p
Design of a SOFC single cell test device
No full textThe electrical performance of single cell of anode-supported planar solid oxide fuel cell (SOFC) received much concern. In this paper, measurement principle, design method and experimental processes of the electrical performance test device for SOFC single cell were introduced. The fluid system of the device consisted of flowmeters, valves, tubes, furnace, etc. The measurement system of the device consisted of digital direct current voltage meter, chromatography work station, and industrial computer, etc. The operator could aquire the test data by the electronic load and work station operation software in the industrial computer
Depth aware RCNN
No full textImage object detection networks that depend on region proposal networks (RPN) have achieved state-of-art results. As RPN is trained to share convolutional features with the actual classification layers in the network, features learned by the convolutional backbones may have subtle impact on the RPN. A successful approach comes from RGB-D image object detection, where the convolutional layers learn not just RGB features, but also depth features.
In this thesis, we study the problem of simultaneously localizing objects as well as estimating their depth. We propose to use one backbone network for two tasks and show that multi-task learning with shared weights can have reciprocating benefits. Our experiments show that when combined with depth prediction in the network, the object detection branch in our model outperforms Faster-RCNN on the challenging KITTI detection benchmark and the Cityscapes dataset. Likewise, the performance of our depth prediction branch is slightly better compared with methods using the same depth prediction architecture.Submission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2021-05-01The student, Tianxi Zhao, accepted the attached license on 2019-04-24 at 08:05.The student, Tianxi Zhao, submitted this Thesis for approval on 2019-04-24 at 08:06.This Thesis was approved for publication on 2019-04-24 at 09:28.DSpace SAF Submission Ingestion Package generated from Vireo submission #13865 on 2019-08-22 at 15:07:58Made available in DSpace on 2019-08-23T20:36:08Z (GMT). No. of bitstreams: 2
ZHAO-THESIS-2019.pdf: 12152227 bytes, checksum: 8f632c204c224a70a8ada07d01502b8f (MD5)
LICENSE.txt: 4208 bytes, checksum: 050255cdbf9658c31eae4895de3316a5 (MD5)
Previous issue date: 2019-04-24Embargo set by: Seth Robbins for item 112201
Lift date: 2021-08-23T20:36:18Z
Reason: Author requested U of Illinois access only (OA after 2yrs) in Vireo ETD systemU of I Only Restriction Lifted for Item 112201 on 2021-08-24T09:15:31Z
Multiple scale sharing faster-RCNN
No full textThe student, Siwei Tang, submitted this Thesis for approval on 2019-04-24 at 10:10.This Thesis was approved for publication on 2019-04-24 at 13:17.DSpace SAF Submission Ingestion Package generated from Vireo submission #13771 on 2019-08-22 at 16:23:20Made available in DSpace on 2019-08-23T20:48:20Z (GMT). No. of bitstreams: 2
TANG-THESIS-2019.pdf: 681796 bytes, checksum: 2c015d2dfaaa358501f9c6f1ab80b395 (MD5)
LICENSE.txt: 4207 bytes, checksum: 59f71a91dfa4f8be3bed256450de5c74 (MD5)
Previous issue date: 2019-04-24Embargo set by: Seth Robbins for item 112352
Lift date: 2021-08-23T20:48:32Z
Reason: Author requested closed access (OA after 2yrs) in Vireo ETD systemLimited Restriction Lifted for Item 112352 on 2021-08-24T09:15:10Z.Small object detection is a challenging task in the field of computer vision because the objects are always of low resolution in the original image and can be easily affected by noise. The state-of-the-art Faster RCNN object detector has good capacity of detecting large objects while small object detection is not one of its advantages. This thesis presents a novel object detector Multi-Scale Sharing Faster-RCNN (MSS-FRCNN) to solve the problem of poor detection performance of small objects by Faster RCNN. We find that upsampling the input image can benefit the small object detection performance. So MSS-FRCNN takes two images with different scales as input and then uses the two feature maps extracted from two images for RoI generation independently. Finally, the model merges the two feature map for classification and bounding box regression. We test our model with two datasets Tsinghua-Tencent 100k and Pascal VOC 07+12. The result demonstrates that MSS-FRCNN can outperform original Faster RCNN in small object detection.Submission published under a 24 month embargo labeled 'Closed Access', the embargo will last until 2021-05-01The student, Siwei Tang, accepted the attached license on 2019-04-24 at 09:11
Effect of ozone addition on oblique detonations in hydrogen-air mixtures
No full textOzone addition is a promising method to enhance the ignition processes and may have potential use in the practical application of oblique detonation engines (ODEs), especially under challenging low Mach number conditions where ignition is hard to achieve. In this study, numerical simulations are conducted to explore the effect of ozone addition on oblique detonations. Numerical results show that with ozone addition, the transition between the induction oblique shock wave (OSW) and oblique detonation wave (ODW) will be smoother at relatively high flight Mach number (from 9 to 10). The reasons for this wave system transformation are dis-cussed. With the addition of ozone, the initiation of ODW can occur at a shorter distance and height, preventing the convergence of compression waves into shocks and resulting in a smoother transition. It was also found that adding ozone is an effective approach to decrease the initiation length, especially at a relatively low Mach number (9), which is crucial to the ODE's thermal protection. Furthermore, the effects of ozone addition on total pressure have been analyzed, demonstrating its positive influence on the propulsion performance of ODE. At a lower flight Mach number (8), ozone addition can re-stabilize the unsteady ODW with a normal detonation wave moving upstream. This is achieved by reducing the size of the subsonic regions, indicating that ozone addition can broaden the working flight Mach numbers of ODEs
Differential treatment for stuff and things: A simple unsupervised domain adaptation method for semantic segmentation
No full textWe consider the problem of unsupervised domain adaptation for semantic segmentation by easing the domain shift between the source domain (synthetic data) and the target domain (real data) in this work. State-of-the-art approaches prove that performing semantic-level alignment is helpful in tackling the domain shift issue. Based on the observation that stuff categories usually share similar appearances across images of different domains while things (i.e. object instances) have much larger differences, we propose to improve the semantic-level alignment with different strategies for stuff regions and for things: (1) for the stuff categories, we generate the feature representation for each class and conduct the alignment operation from the target domain to the source domain; (2) for the thing categories, we generate the feature representation for each individual instance and encourage the instance in the target domain to align with the most similar one in the source domain. In this way, the individual differences within thing categories will also be considered to alleviate over-alignment. In addition to our proposed method, we further reveal the reason why the current adversarial loss is often unstable in minimizing the distribution discrepancy and show that our method can help ease this issue by minimizing the most similar stuff and instance features between the source and the target domains. We conduct extensive experiments in two unsupervised domain adaptation tasks, GTA5 to Cityscapes and SYNTHIA to Cityscapes, and achieve the new state-of-the-art segmentation accuracy.Submission published under a 24 month embargo labeled 'Closed Access', the embargo will last until 2021-12-01The student, Zhonghao Wang, accepted the attached license on 2019-11-22 at 14:13.The student, Zhonghao Wang, submitted this Thesis for approval on 2019-11-22 at 14:25.This Thesis was approved for publication on 2019-11-22 at 15:53.DSpace SAF Submission Ingestion Package generated from Vireo submission #14591 on 2020-02-28 at 17:36:44Made available in DSpace on 2020-03-02T22:38:46Z (GMT). No. of bitstreams: 2
WANG-THESIS-2019.pdf: 1057263 bytes, checksum: c4adea20e1084d978a416296afffba84 (MD5)
LICENSE.txt: 4210 bytes, checksum: c435e8b7ce1ce7c1d89116552470a39f (MD5)
Previous issue date: 2019-11-22Embargo set by: Seth Robbins for item 114001
Lift date: 2022-03-02T22:39:04Z
Reason: Author requested closed access (OA after 2yrs) in Vireo ETD systemLimited Restriction Lifted for Item 114001 on 2022-03-03T10:15:27Z
Exosomes-Based Nanotherapeutic Strategies: An Important Approach for Spinal Cord Injury Repair
No full textCheng Ju,1,2,&ast; Hui Dong,1,2,&ast; Renfeng Liu,1,2,&ast; Xuan Wang,1,2 Ruiqing Xu,1,2 Huimin Hu,1,2 Dingjun Hao1,2 1Department of Spine Surgery, Honghui Hospital, Xi’an Jiaotong University, Xi’an, Shaanxi, People’s Republic of China; 2Shaanxi Key Laboratory of Spine Bionic Treatment, Honghui Hospital, Xi’an Jiaotong University, Xi’an, Shaanxi, People’s Republic of China&ast;These authors contributed equally to this workCorrespondence: Dingjun Hao, Email [email protected] Huimin Hu, Email [email protected]: The repair and functional regeneration of spinal cord injury (SCI) remains a major challenge and focal point in regenerative medicine. Following SCI significant inflammation and neuronal damage occur. Conventional drug therapies often fail to precisely target the injured areas and cannot cross the blood-spinal cord barrier, severely limiting therapeutic efficacy. Therefore, precision therapeutics are crucial to improve the prognosis of SCI patients. In recent years, exosomes have gained widespread attention as natural delivery vehicles due to their low immunogenicity, high biocompatibility, and efficient delivery capabilities. Exosomes can effectively cross cell membranes and target specific cells, playing an important role in intercellular signaling. This makes them highly promising for precision therapies in SCI. By engineering exosomes for targeted delivery, new strategies can be developed for drug delivery, gene therapy, and personalized treatment after SCI. We aimed to review the biological functions of exosomes derived from different cell sources and discuss the role in tissue repair following SCI. Additionally, we explore the prospects and potential of exosomes in clinical SCI applications, to provide valuable research insights to improve functional recovery and long-term health management for SCI patients in the future.Keywords: spinal cord injury, exosomes, nanodelivery, inflammation, nerve injury, non-coding RN
- …
