A5040475616- Organic Light-Emitting Diodes Research
- Organic Electronics and Photovoltaics
- Nerve injury and regeneration
- Conducting polymers and applications
- Face recognition and analysis
- Angiogenesis and VEGF in Cancer
- Tuberculosis Research and Epidemiology
- Face and Expression Recognition
- Pharmacological Receptor Mechanisms and Effects
- Luminescence and Fluorescent Materials
- Mycobacterium research and diagnosis
- Phenothiazines and Benzothiazines Synthesis and Activities
- Mesenchymal stem cell research
- Bacteriophages and microbial interactions
- Axon Guidance and Neuronal Signaling
- Neurogenesis and neuroplasticity mechanisms
- Receptor Mechanisms and Signaling
- Electrospun Nanofibers in Biomedical Applications
- Soil Carbon and Nitrogen Dynamics
- Bluetooth and Wireless Communication Technologies
- Echinoderm biology and ecology
- Advanced Algorithms and Applications
- Crystallization and Solubility Studies
- Graphene and Nanomaterials Applications
- Tissue Engineering and Regenerative Medicine
Jiaying University
2025
Jilin University
2025
Affiliated Hospital of North Sichuan Medical College
2024-2025
Changzhou University
2025
Xuzhou Medical College
2020-2024
Computer Network Information Center
2019-2024
University of Ottawa
2012-2024
China University of Petroleum, Beijing
2010-2024
University of Chinese Academy of Sciences
2019-2024
China Spallation Neutron Source
2024
This paper presents a highly efficient, very accurate regression approach for face alignment. Our has two novel components: set of local binary features, and locality principle learning those features. The guides us to learn discriminative features each facial landmark independently. obtained are used jointly linear the final output. achieves state-of-the-art results when tested on current most challenging benchmarks. Furthermore, because extracting regressing is computationally cheap, our...
Making a high-dimensional (e.g., 100K-dim) feature for face recognition seems not good idea because it will bring difficulties on consequent training, computation, and storage. This prevents further exploration of the use high dimensional feature. In this paper, we study performance We first empirically show that dimensionality is critical to performance. A 100K-dim feature, based single-type Local Binary Pattern (LBP) descriptor, can achieve significant improvements over both its...
We present a very efficient, highly accurate, "Explicit Shape Regression" approach for face alignment. Unlike previous regression-based approaches, we directly learn vectorial regression function to infer the whole facial shape (a set of landmarks) from image and explicitly minimize alignment errors over training data. The inherent constraint is naturally encoded into regressor in cascaded learning framework applied coarse fine during test, without using fixed parametric model as most...
We study the problem of automatically removing outliers from noisy data, with application for outlier images an image collection. address this by utilizing reconstruction errors autoencoder. observe that when data are reconstructed low-dimensional representations, inliers and can be well separated according to their errors. Based on basic observation, we gradually inject discriminative information in learning process autoencoder make more separable. Experiments a variety datasets validate...
Recently, deep learning approaches have demonstrated remarkable progresses for action recognition in videos. Most existing frameworks equally treat every volume i.e. spatial-temporal video clip, and directly assign a label to all volumes sampled from it. However, within video, discriminative actions may occur sparsely few key volumes, most other are irrelevant the labeled category. Training with large proportion of will hurt performance. To address this issue, we propose mining framework...
The analysis of the behavior soil and foundations when piles in offshore areas are subjected to long-term lateral loading (wind) is one major problems associated with smooth operation superstructure. strength pile-soil system influenced by variations water content soil. At present, there no studies carried out analyzing mechanical deformational both material laterally loaded groundwater level as a variable. In this paper, series 1-g model tests were conducted explore monopile under...
Phase noise in mutually synchronized oscillator systems is analyzed for arbitrary coupling and injection-locking topologies, neglecting amplitude noise, modulation (AM) to phase (PM) conversion. When the chosen properly (depending on model), near-carrier reduced 1/N that of a single oscillator, provided network reciprocal. This proved In general, illustrated with specific cases globally coupled nearest-neighbor chains. A slight degradation found unilaterally (nonreciprocal) The reduction...
Face verification involves determining whether a pair of facial images belongs to the same or different subjects. This problem can prove be quite challenging in many important applications where labeled training data is scarce, e.g., family album photo organization software. Herein we propose principled transfer learning approach for merging plentiful source-domain with limited samples from some target domain interest create classifier that ideally performs nearly as well if rich...
Cyanopyridine is used as a new electron-withdrawing unit for thermally activated delayed fluorescence (TADF). Solution-processed simple structure devices using 2,3,5,6-tetracarbazole-4-cyano-pyridine green TADF dopant show maximum external quantum efficiencies high 10.4% and 11.4% with 2,3,5,6-tetracarbazole-pyridine 1,3-bis(N-carbazolyl)benzene host materials, respectively.
This paper presents a highly efficient and accurate regression approach for face alignment. Our has two novel components: 1) set of local binary features 2) locality principle learning those features. The guides us to learn discriminative each facial landmark independently. obtained are used jointly linear the final output. achieves state-of-the-art results when tested on most challenging benchmarks date. Furthermore, because extracting regressing computationally very cheap, our system is...
Non-conjugated trifluoromethyl was introduced as a new acceptor in TADF OLEDs. 5CzCF<sub>3</sub>Ph exhibited significantly higher efficiency than 4CzCF<sub>3</sub>Ph due to its lower Δ<italic>E</italic><sub>ST</sub> value.
We describe a catalytic strategy for direct single C(sp3)-F bond alkylation of trifluoromethylbenzimidazoles under photoinduced thiol catalysis process. The CO2 radical anion (CO2•-) proved to be the most efficient single-electron reductant realize such transformation. spin-center shift generated intermediate is key step in realizing C-F activation mild conditions with high efficiency.
Extending the π-skeletons of multi-resonance (MR) organoboron emitters can feasibly modulate their optoelectronic properties. Here, we first adopt indolo[3,2-b]indole (32bID) segment as a multi-nitrogen bridge and develop high-efficiency π-extended narrowband green emitter. This moiety establishes not only high-yield one-shot multiple Bora-Friedel-Crafts reaction towards MR skeleton, but compact N-ethylene-N motif for red-shifted emission. An emission peak at 524 nm, small full width half...
Gut microbiota dysbiosis significantly impacts ulcerative colitis (UC) progression and exacerbation. Probiotics show promise in UC management. This study evaluated the effects of different doses Bacillus pumilus LV149, an aquatic-derived probiotic, on gut injury repair male C57BL/6 mice with dextran sulfate sodium (DSS)-induced investigated underlying mechanisms. was induced by allowing free access to a 3% DSS solution for 7 days, concurrent daily oral gavage either low (LV149-L, 1 × 108...
Random forest is well known as one of the best learning methods. In spite its great success, it also has certain drawbacks: heuristic rule does not effectively minimize global training loss; model size usually too large for many real applications. To address issues, we propose two techniques, refinement and pruning, to improve a pre-trained random forest. The proposed jointly relearns leaf nodes all trees under objective function so that complementary information between multiple exploited....