Learning From Crowds With Multiple Noisy Label Distribution Propagation

Author(s):  
Liangxiao Jiang ◽  
Hao Zhang ◽  
Fangna Tao ◽  
Chaoqun Li
Keyword(s):  
Author(s):  
Richard W. Burry ◽  
Diane M. Hayes

Electron microscopic (EM) immunocytochemistry localization of the neuron specific protein p65 could show which organelles contain this antigen. Antibodies (Ab) labeled with horseradish peroxidase (HRP) followed by chromogen development show a broad diffuse label distribution within cells and restricting identification of organelles. Particulate label (e.g. 10 nm colloidal gold) is highly desirable but not practical because penetration into cells requires destroying the plasma membrane. We report pre-embedding immunocytochemistry with a particulate marker, 1 nm gold, that will pass through membranes treated with saponin, a mild detergent.Cell cultures of the rat cerebellum were fixed in buffered 4% paraformaldehyde and 0.1% glutaraldehyde (Glut.). The buffer for all incubations and rinses was phosphate buffered saline with: 1% calf serum, 0.2% saponin, 0.1% gelatin, 50 mM glycine 1 mg/ml bovine serum albumin, and (not in the HRP labeled cultures) 0.02% sodium azide. The monoclonal #48 to p65 was used with three label systems: HRP, 1 nm avidin gold with IntenSE M development, and 1 nm avidin gold with Danscher development.


2021 ◽  
Vol 436 ◽  
pp. 12-21
Author(s):  
Xinyue Dong ◽  
Shilin Gu ◽  
Wenzhang Zhuge ◽  
Tingjin Luo ◽  
Chenping Hou

1984 ◽  
Vol 62 (6) ◽  
pp. 1149-1157 ◽  
Author(s):  
D. Driss-Ecole ◽  
G. Perbal ◽  
Y. Leroux

[3H]indoleacetic acid (AIA) was applied to the shoot tip of intact young plants of Tomato (Lycopersicum esculentum Mill.) for 10, 60, or 120 min. Autoradiograms of whole plants were prepared and liquid scintillation counts of stem segments and principal root segments were performed. Chromatographic analysis showed that 66% of the radioactivity was associated with AIA after 120 min of contact with [3H]AIA. Autoradiographs of semithin and ultrathin sections were prepared after treatment by DCC (1-(3-dimethyl-aminopropyl)-3-ethylcarbodiimide hydrochloride). The quantity of label per cell and the density of label were determined for all tissues of the apical bud. The density of label was greater for meristematic cells than for differentiated cells. The observed homogeneity of label distribution in the apical meristem shows that auxin levels do not play a prominent role in distinguishing between its lateral and axial zones. The density of label was similar in the apical and in the axillary bud of leaf 4. The cells of the rib meristem, which elongate to produce pith, were more intensely labelled than the other meristematic cells. The percentage of label was calculated for each tissue in a transverse section of the stem just below the apex. The amount of auxin was greatest in the parenchyma (axial and cortical) with lesser amounts in the procambium, phloem parenchyma, and xylem parenchyma. Vessels, which had the greatest density of label, did not contain more than about 3% of total radioactivity of the stem section, while sieve tubes had only 0.5%. Pathways of auxin transport and the role of AIA in regulating meristematic activity in the apical bud are discussed.


Author(s):  
Xiuyi Jia ◽  
Xiaoxia Shen ◽  
Weiwei Li ◽  
Yunan Lu ◽  
Jihua Zhu

Author(s):  
Hao Zhang ◽  
Liangxiao Jiang ◽  
Wenqiang Xu

Crowdsourcing services provide a fast, efficient, and cost-effective means of obtaining large labeled data for supervised learning. Ground truth inference, also called label integration, designs proper aggregation strategies to infer the unknown true label of each instance from the multiple noisy label set provided by ordinary crowd workers. However, to the best of our knowledge, nearly all existing label integration methods focus solely on the multiple noisy label set itself of the individual instance while totally ignoring the intercorrelation among multiple noisy label sets of different instances. To solve this problem, a multiple noisy label distribution propagation (MNLDP) method is proposed in this study. MNLDP first transforms the multiple noisy label set of each instance into its multiple noisy label distribution and then propagates its multiple noisy label distribution to its nearest neighbors. Consequently, each instance absorbs a fraction of the multiple noisy label distributions from its nearest neighbors and yet simultaneously maintains a fraction of its own original multiple noisy label distribution. Promising experimental results on simulated and real-world datasets validate the effectiveness of our proposed method.


Author(s):  
Yongbiao Gao ◽  
Yu Zhang ◽  
Xin Geng

Label distribution learning (LDL) is a novel machine learning paradigm that gives a description degree of each label to an instance. However, most of training datasets only contain simple logical labels rather than label distributions due to the difficulty of obtaining the label distributions directly. We propose to use the prior knowledge to recover the label distributions. The process of recovering the label distributions from the logical labels is called label enhancement. In this paper, we formulate the label enhancement as a dynamic decision process. Thus, the label distribution is adjusted by a series of actions conducted by a reinforcement learning agent according to sequential state representations. The target state is defined by the prior knowledge. Experimental results show that the proposed approach outperforms the state-of-the-art methods in both age estimation and image emotion recognition.


Sign in / Sign up

Export Citation Format

Share Document