Bayesian losses for homoscedastic aleatoric uncertainty modeling in pollen image detection
Annotation
The paper investigates the homoscedastic aleatoric uncertainty modeling for the detection of pollen in images. The new uncertainty modeling loss functions are presented, which are based on the focal and smooth L1 losses. The focal and smooth L1 losses proved their efficiency for the problem of image detection, however, they do not allow modeling the aleatoric uncertainty, while the proposed functions do, leading to more accurate solutions. The functions are based on Bayesian inference and allow for effortless use in existing neural network detectors based on the RetinaNet architecture. The advantages of the loss functions are described on the problem of pollen detection in images. The new loss functions increased the accuracy of pollen image detection, namely localization and classification, on average by 2.76 %, which is crucial for the pollen recognition in general. This helps to automate the process of determining allergenic pollen in the air and reduce the time to inform patients with pollinosis to prevent allergy symptoms. The obtained result shows that the modeling of homoscedastic aleatoric uncertainty for neural networks allows separating the noise from the data, increasing the accuracy of the proposed solutions. The developed functions can be applied to train neural network detectors on any other image datasets.
Keywords
Постоянный URL
Articles in current issue
- On the feasibility of the monostatic scheme for constructing the land-based telescope at supervision of space objects
- DREM procedure application for piecewise constant parameters identification
- Features of the morphology of micro- and nanoporous copper and silver films synthesized by substitution reaction for photocatalytic application.
- Nature-inspired metaheuristic scheduling algorithms in cloud: a systematic review
- Evaluation of the applicability of asynchronous programming methods to the data consistency problem in a microservices environment
- A factor model for detection and recognition of human face contours and elements
- A study of the stability of information and telecommunication networks under conditions of stochastic percolation of nodes
- Context-free path querying with all-path semantics using matrices with sets of intermediate vertices.
- Decision support system for the proton therapy implementation
- Determination of dangerous driving behavior based on the use of information from wearable electronic devices
- An automata-based programming engine
- The speech synthesis detection algorithm based on cepstral coefficients and convolutional neural network
- Risk assessment methodology for information systems, based on the user behavior and IT-security incidents analysis
- Identification of user accounts by image comparison: the pHash-based approach
- A study of human motion in computer vision systems based on a skeletal model
- Solution of super- and hypersonic gas dynamic problems with a model of high-temperature air
- Modeling security violation processes in machine learning systems
- Mathematical modeling of an optimal oncotherapy for malignant tumors.
- A numerical study of the expansion of a gas-particles mixture with axial symmetry.
- The study of the birefrigence modulator based on lithium niobate