| 1. |
- Kristan, Matej, et al.
(författare)
-
The Visual Object Tracking VOT2015 challenge results
- 2015
-
Ingår i: Proceedings 2015 IEEE International Conference on Computer Vision Workshops ICCVW 2015. - : IEEE. - 9780769557205 ; , s. 564-586
-
Konferensbidrag (refereegranskat)abstract
- The Visual Object Tracking challenge 2015, VOT2015, aims at comparing short-term single-object visual trackers that do not apply pre-learned models of object appearance. Results of 62 trackers are presented. The number of tested trackers makes VOT 2015 the largest benchmark on short-term tracking to date. For each participating tracker, a short description is provided in the appendix. Features of the VOT2015 challenge that go beyond its VOT2014 predecessor are: (i) a new VOT2015 dataset twice as large as in VOT2014 with full annotation of targets by rotated bounding boxes and per-frame attribute, (ii) extensions of the VOT2014 evaluation methodology by introduction of a new performance measure. The dataset, the evaluation kit as well as the results are publicly available at the challenge website(1).
|
|
| 2. |
- Kristan, Matej, et al.
(författare)
-
The Visual Object Tracking VOT2016 Challenge Results
- 2016
-
Ingår i: COMPUTER VISION - ECCV 2016 WORKSHOPS, PT II. - Cham : SPRINGER INT PUBLISHING AG. - 9783319488813 - 9783319488806 ; , s. 777-823
-
Konferensbidrag (refereegranskat)abstract
- The Visual Object Tracking challenge VOT2016 aims at comparing short-term single-object visual trackers that do not apply pre-learned models of object appearance. Results of 70 trackers are presented, with a large number of trackers being published at major computer vision conferences and journals in the recent years. The number of tested state-of-the-art trackers makes the VOT 2016 the largest and most challenging benchmark on short-term tracking to date. For each participating tracker, a short description is provided in the Appendix. The VOT2016 goes beyond its predecessors by (i) introducing a new semi-automatic ground truth bounding box annotation methodology and (ii) extending the evaluation system with the no-reset experiment.
|
|
| 3. |
- Chen, Mingming, et al.
(författare)
-
Carbon anode in direct carbon fuel cell
- 2010
-
Ingår i: International journal of hydrogen energy. - : Elsevier BV. - 0360-3199 .- 1879-3487. ; 35:7, s. 2732-2736
-
Tidskriftsartikel (refereegranskat)abstract
- Direct carbon fuel cell (DCFC) is a kind of high temperature fuel cell using carbon materials directly as anode. Electrochemical reactivity and surface property of carbon were taken into account in this paper. Four representative carbon samples were selected. The most suitable ratio of the ternary eutectic mixture Li2CO3-K2CO3-Al2O3 was determined at 1.05:1.2:1(mass ration). Conceptual analysis for electrochemical reactivity of carbon anode shows the importance of (1) reactive characteristics including lattice disorder, edge-carbon ratio and the number of short alkyl side chain of carbon material, which builds the prime foundation of the anodic half-cell reaction; (2) surface wetting ability, which assures the efficient contact of anode surface with electrolyte. It indicates that anode reaction rate and DCFC output can be notably improved if carbon are pre-dispersed into electrolyte before acting as anode, due to the straightway shift from cathode to anode for CO32- provided by electrolyte soaked in carbon material.
|
|
| 4. |
- Liu, Qiong, et al.
(författare)
-
Vertical and horizontal shifts in the microbial community structure of paddy soil under long-term fertilization regimes
- 2022
-
Ingår i: Applied Soil Ecology. - : Elsevier BV. - 0929-1393. ; 169
-
Tidskriftsartikel (refereegranskat)abstract
- Knowledge remains limited on how the structure of microbial community in paddy soils changes in relation to different types of fertilizers with same amount of nutrients. Thus, here, soil samples were collected at 0–10, 10–20, 20–30, and 30–40 cm depths from a paddy field subjected to four long-term fertilization treatments (no fertilization, mineral fertilization, mineral fertilization combined with rice straw, and chicken manure) and analyzed for microbial biomass and community composition. In unfertilized soils, microbial biomass decreased from 0 to 40 cm (with actinomycetes < gram-positive (G+) bacteria < gram-negative (G? ) bacteria < fungi). This ordering was retained after fertilization, but the decline with depth was less pronounced. Both mineral and mineral plus organic fertilization increased the biomass of G+ bacteria compared to G? bacteria (22.7–56.2% increase) and actinomycetes (14.8–52.5% increase). Thus, over the long term, G+ bacteria benefited the most from mineral fertilizer than the other microbial groups. The partial replacement of mineral fertilizer with manure primarily enhanced the abundance of G+ bacteria at 0–30 cm soil depth, whereas replacement with straw enhanced the abundance of fungi at 10–20 cm soil depth. Our findings demonstrate that the structure of the microbial community is strongly impacted by long-term fertilization, independent of fertilizer type.
|
|
| 5. |
- Liu, Yuhuai, et al.
(författare)
-
Stoichiometric theory shapes enzyme kinetics in paddy bulk soil but not in rhizosphere soil
- 2022
-
Ingår i: Land Degradation and Development. - : Wiley. - 1099-145X .- 1085-3278. ; 33:2, s. 246-256
-
Tidskriftsartikel (refereegranskat)abstract
- The available carbon (C) to phosphorus (P) ratio in soil is regulated by extracellular hydrolases for C and P acquisition by microbes and plants. However, the stoichiometric relationship between acquiring C and P in paddy rhizosphere and bulk soils remains unclear. The objective was to explore the underlying mechanisms of C and P acquisition stoichiometry in rhizosphere and bulk soils in response to P fertilization and cellulose addition. Amendment with either cellulose or P separately caused a significant increase in the maximal velocity (Vmax) of C acquisition enzymes (β-1,4-glucosidase and β-cellobiohydrolase) but decreased that of P acquisition enzymes (acid and alkaline phosphomonoesterases) in bulk soil. In contrast, lower Vmax values of C and P acquisition enzymes were observed in rhizosphere soil than in bulk soil. The co-application of cellulose and P increased the Vmax of P acquisition enzymes in rhizosphere soil but decreased that of only alkaline phosphomonoesterase in bulk soil. Results show that P availability and labile-C content co-regulated the P/C acquisition ratio, and two inverse linear relationships were observed. Specifically, the P/C acquisition ratio was negatively related to both the dissolved organic C/Olsen-P ratio and the microbial biomass C/P ratio in rhizosphere soil. However, the P/C acquisition ratio was positively related to both the dissolved organic C/Olsen-P ratio and the microbial biomass C/P ratio in bulk soil. Overall, microbes mineralized less organic P to acquire P in paddy soil rhizosphere (i.e. containing higher labile-C) than in bulk soil (i.e. having lower labile-C contents).
|
|
| 6. |
- Posey, Victoria A., et al.
(författare)
-
Two-dimensional heavy fermions in the van der Waals metal CeSiI
- 2024
-
Ingår i: Nature. - : Springer Nature. - 0028-0836 .- 1476-4687. ; 625:7995, s. 483-488
-
Tidskriftsartikel (refereegranskat)abstract
- Heavy-fermion metals are prototype systems for observing emergent quantum phases driven by electronic interactions1-6. A long-standing aspiration is the dimensional reduction of these materials to exert control over their quantum phases7-11, which remains a significant challenge because traditional intermetallic heavy-fermion compounds have three-dimensional atomic and electronic structures. Here we report comprehensive thermodynamic and spectroscopic evidence of an antiferromagnetically ordered heavy-fermion ground state in CeSiI, an intermetallic comprising two-dimensional (2D) metallic sheets held together by weak interlayer van der Waals (vdW) interactions. Owing to its vdW nature, CeSiI has a quasi-2D electronic structure, and we can control its physical dimension through exfoliation. The emergence of coherent hybridization of f and conduction electrons at low temperature is supported by the temperature evolution of angle-resolved photoemission and scanning tunnelling spectra near the Fermi level and by heat capacity measurements. Electrical transport measurements on few-layer flakes reveal heavy-fermion behaviour and magnetic order down to the ultra-thin regime. Our work establishes CeSiI and related materials as a unique platform for studying dimensionally confined heavy fermions in bulk crystals and employing 2D device fabrication techniques and vdW heterostructures12 to manipulate the interplay between Kondo screening, magnetic order and proximity effects.
|
|
| 7. |
- Xu, Sihong, et al.
(författare)
-
Carbon doped MO-SDC material as an SOFC anode
- 2007
-
Ingår i: Journal of Power Sources. - : Elsevier BV. - 0378-7753 .- 1873-2755. ; 165:1, s. 82-86
-
Tidskriftsartikel (refereegranskat)abstract
- Oxide mixtures MO-SDC, M=Cu, Ni, Co, SDC=Ce0.9Sm0.1O1.95 were synthesized by employing a citrate/nitrate combustion technique. Two kinds of Carbon materials, activated carbon (AC) and vapor grown carbon fiber (VGCF) were homogeneously dispersed into the MO-SDC. The materials can be used as anodes to fabricate single cells using a uniaxial die-press method. The sintering temperature was studied to optimize cell performance. Experimental results showed that cells sintered at 700 degrees C had better performance. When the temperature was above 750 degrees C, the cells were severely distorted, and cannot be tested. Compared with the basic MO-SDC anode, AC and VGCF improve the solid oxide fuel cell (SOFC) anode properties, due to a change of the microstructures of the anode materials which enhance their electron conductivity. Single cell performances were evaluated by I-V measurements, and when 1.25 wt.%VGCF was introduced into the MO-SDC by ball-milling, termed: 1.25 wt.%VGCF-MO-SDC, the 1.25 wt.%VGCF-MO-SDC anode material could achieve the highest power density of up to 0.326 W cm(-2) with H-2 as fuel. The calcination temperature of the MO-SDC dry gel also strongly influenced the electrochemical performance of the 1.25 wt.%VGCF-MO-SDC material. XRD spectra for each calcined temperature and the I-V measurement both suggest that calcinations at 550 degrees C for 1 h are suitable. 1.0 wt.%AC-MO-SDC and 1.25 wt.%VGCF-MO-SDC have similar performance when the cell was fed in methanol/3%H2O, and the corresponding power density was up to 0.253 W cm(-2). Traces of carbon were found in the off-gases.
|
|
| 8. |
- Zhu, Xiaomeng, et al.
(författare)
-
Automated assembly quality inspection by deep learning with 2D and 3D synthetic CAD data
- 2024
-
Ingår i: Journal of Intelligent Manufacturing. - : Springer Nature. - 0956-5515 .- 1572-8145.
-
Tidskriftsartikel (refereegranskat)abstract
- In the manufacturing industry, automatic quality inspections can lead to improved product quality and productivity. Deep learning-based computer vision technologies, with their superior performance in many applications, can be a possible solution for automatic quality inspections. However, collecting a large amount of annotated training data for deep learning is expensive and time-consuming, especially for processes involving various products and human activities such as assembly. To address this challenge, we propose a method for automated assembly quality inspection using synthetic data generated from computer-aided design (CAD) models. The method involves two steps: automatic data generation and model implementation. In the first step, we generate synthetic data in two formats: two-dimensional (2D) images and three-dimensional (3D) point clouds. In the second step, we apply different state-of-the-art deep learning approaches to the data for quality inspection, including unsupervised domain adaptation, i.e., a method of adapting models across different data distributions, and transfer learning, which transfers knowledge between related tasks. We evaluate the methods in a case study of pedal car front-wheel assembly quality inspection to identify the possible optimal approach for assembly quality inspection. Our results show that the method using Transfer Learning on 2D synthetic images achieves superior performance compared with others. Specifically, it attained 95% accuracy through fine-tuning with only five annotated real images per class. With promising results, our method may be suggested for other similar quality inspection use cases. By utilizing synthetic CAD data, our method reduces the need for manual data collection and annotation. Furthermore, our method performs well on test data with different backgrounds, making it suitable for different manufacturing environments.
|
|
| 9. |
- Zhu, Xiaomeng, et al.
(författare)
-
Automatic assembly quality inspection based on an unsupervised point cloud domain adaptation model
- 2021
-
Ingår i: Procedia CIRP. - : Elsevier BV. - 2212-8271. ; , s. 1801-1806
-
Konferensbidrag (refereegranskat)abstract
- This paper proposes an end-to-end method for automatic assembly quality inspection based on a point cloud domain adaptation model. The method involves automatically generating labeled point clouds from various CAD models and training a model on those point clouds together with a limited number of unlabeled point clouds acquired by 3D cameras. The model can then classify newly captured point clouds from 3D cameras to execute assembly quality inspection with promising performance. The method has been evaluated in an industry case study of pedal car front-wheel assembly. By utilizing CAD data, the method is less time-consuming for implementation in production.
|
|
| 10. |
- Zhu, Xiaomeng, et al.
(författare)
-
Surface Defect Detection with Limited Training Data : A Case Study on Crown Wheel Surface Inspection
- 2023
-
Ingår i: Procedia CIRP. - : Elsevier. - 2212-8271 .- 2212-8271. ; 120, s. 1333-1338, s. 1333-1338
-
Tidskriftsartikel (refereegranskat)abstract
- This paper presents an approach to automatic surface defect detection by a deep learning-based object detection method, particularly in challenging scenarios where defects are rare, i.e., with limited training data. We base our approach on an object detection model YOLOv8, preceded by a few steps: 1) filtering out irrelevant information, 2) enhancing the visibility of defects, namely brightness contrast, and 3) increasing the diversity of the training data through data augmentation. We evaluated the method in an industrial case study of crown wheel surface inspection in detecting Unclean Gear as well as Deburring defects, resulting in promising performances. With the combination of the three preprocessing steps, we improved the detection accuracy by 22.2% and 37.5% respectively while detecting those two defects. We believe that the proposed approach is also adaptable to various applications of surface defect detection in other industrial environments as the employed techniques, such as image segmentation, are available off the shelf.
|
|
| 11. |
- Zhu, Xiaomeng, et al.
(författare)
-
Towards Sim-to-Real Industrial Parts Classification with Synthetic Dataset
- 2023
-
Ingår i: Proceedings, 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops. - : IEEE. - 9798350302493 - 9798350302509 ; , s. 4454-4463, s. 4454-4463
-
Konferensbidrag (refereegranskat)abstract
- This paper is about effectively utilizing synthetic data for training deep neural networks for industrial parts classification, in particular, by taking into account the domain gap against real-world images. To this end, we introduce a synthetic dataset that may serve as a preliminary testbed for the Sim-to-Real challenge; it contains 17 objects of six industrial use cases, including isolated and assembled parts. A few subsets of objects exhibit large similarities in shape and albedo for reflecting challenging cases of industrial parts. All the sample images come with and without random backgrounds and post-processing for evaluating the importance of domain randomization. We call it Synthetic Industrial Parts dataset (SIP-17). We study the usefulness of SIP-17 through benchmarking the performance of five state-of-the-art deep network models, supervised and self-supervised, trained only on the synthetic data while testing them on real data. By analyzing the results, we deduce some insights on the feasibility and challenges of using synthetic data for industrial parts classification and for further developing larger-scale synthetic datasets. Our dataset † and code ‡ are publicly available.
|
|
| 12. |
- Zhu, Xiaomeng, et al.
(författare)
-
Unsupervised domain adaptive object detection for assembly quality inspection
- 2022
-
Ingår i: Proceedings 15th CIRP Conference on Intelligent Computation in Manufacturing Engineering, ICME 2021. - : Elsevier BV. ; , s. 477-482
-
Konferensbidrag (refereegranskat)abstract
- A challenge to apply deep learning-based computer vision technologies for assembly quality inspection lies in the diverse assembly approaches and the restricted annotated training data. This paper describes a method for overcoming the challenge by training an unsupervised domain adaptive object detection model on annotated synthetic images generated from CAD models and unannotated images captured from cameras. On a case study of pedal car front-wheel assembly, the model achieves promising results compared to other state-of-the-art object detection methods. Besides, the method is efficient to implement in production as it does not require manually annotated data.
|
|
