Multi-task Approach Research Articles

A multi-task network approach for calculating discrimination-free insurance prices

In applications of predictive modeling, such as insurance pricing, indirect or proxy discrimination is an issue of major concern. Namely, there exists the possibility that protected policyholder characteristics are implicitly inferred from non-protected ones by predictive models and are thus having an undesirable (and possibly illegal) impact on prices. A technical solution to this problem relies on building a best-estimate model using all policyholder characteristics (including protected ones) and then averaging out the protected characteristics for calculating individual prices. However, such an approach requires full knowledge of policyholders’ protected characteristics, which may in itself be problematic. Here, we address this issue by using a multi-task neural network architecture for claim predictions, which can be trained using only partial information on protected characteristics and produces prices that are free from proxy discrimination. We demonstrate the proposed method on both synthetic data and a real-world motor claims dataset, in which proxy discrimination can be observed. In both examples we find that the predictive accuracy of the multi-task network is comparable to a conventional feed-forward neural network, when the protected information is available for at least half of the insurance policies. However, the multi-task network has superior performance in the case when the protected information is known for less than half of the insurance policyholders.

A Comparison of Two Multi-Tasking Approaches to Cognitive Training in Cardiac Surgery Patients.

The multi-tasking approach may be promising for cognitive rehabilitation in cardiac surgery patients due to a significant effect on attentional and executive functions. This study aimed to compare the neuropsychological changes in patients who have undergone two variants of multi-tasking training and a control group in the early postoperative period of coronary artery bypass grafting (CABG). One hundred and ten CABG patients were divided into three groups: cognitive training (CT) I (a postural balance task with mental arithmetic, verbal fluency, and divergent tasks) (n = 30), CT II (a simple visual-motor reaction with mental arithmetic, verbal fluency, and divergent tasks) (n = 40), and control (n = 40). Two or more cognitive indicators improved in 93.3% of CT I patients, in 72.5% of CT II patients, and in 62.5% of control patients; CT I patients differed from CT II and control (p = 0.04 and p = 0.008, respectively). The improving short-term memory and attention was found more frequently in the CT I group as compared to control (56.7% vs. 15%; p = 0.0005). The cognitive improvement of all domains (psychomotor and executive functions, attention, and short-term memory) was also revealed in CT I patients more frequently than CT II (46.7% vs. 20%; p = 0.02) and control (46.7% vs. 5%; p = 0.0005). The CT I multi-tasking training was more effective at improving the cognitive performance in cardiac surgery patients as compared to CT II training and standard post-surgery management. The findings of this study will be helpful for future studies involving multi-tasking training.

Accurate and generalizable photovoltaic panel segmentation using deep learning for imbalanced datasets

The widespread adoption of photovoltaic (PV) technology for renewable energy necessitates accurate segmentation of PV panels to estimate installation capacity. However, achieving highly efficient and precise segmentation methods remains a pressing challenge. Recent advancements in artificial intelligence and remote sensing techniques have shown promise in PV segmentation. Nevertheless, real-world scenarios introduce complexities such as diverse sensing platforms, sensors, panel categories, and testing regions. These factors contribute to resolution, size, and foreground-background class imbalances, impeding accurate and generalized PV panel segmentation over large areas. To address these challenges, we propose GenPV, a deep learning model that leverages data distribution analysis and PV panel characteristics to enhance segmentation accuracy and generalization. GenPV employs a multi-scale feature learning approach, utilizing an enhanced feature pyramid network to fuse data features from multiple resolutions, effectively addressing resolution imbalance. Moreover, inductive learning is employed through a multitask approach, facilitating the detection and identification of both small and large-sized PV panels to mitigate size imbalance. To address significant class imbalance in PV panel recognition tasks, we integrate the Focal loss function for effective hard sample mining. Through experimental evaluation conducted in Heilbronn, Germany, our proposed method demonstrates superior performance compared to state-of-the-art approaches in PV panel segmentation. The results exhibit progressively higher accuracy and improved generalization capability. These findings highlight the potential of our method to serve as an advanced and practical tool for PV segmentation in the renewable energy field.

DC-AAE: Dual channel adversarial autoencoder with multitask learning for KL-grade classification in knee radiographs

Knee osteoarthritis (OA) is a frequent musculoskeletal disorder that leads to physical disability in older adults. Manual OA assessment is performed via visual inspection, which is highly subjective as it suffers from moderate to high inter-observer variability. Many deep learning-based techniques have been proposed to address this issue. However, owing to the limited amount of labelled data, all existing solutions have limitations in terms of performance or the number of classes. This paper proposes a novel fully automatic Kellgren and Lawrence (KL) grade classification scheme in knee radiographs. We developed a semi-supervised multi-task learning-based approach that enables the exploitation of additional unlabelled data in an unsupervised as well as supervised manner. Specifically, we propose a dual-channel adversarial autoencoder, which is first trained in an unsupervised manner for reconstruction tasks only. To exploit the additional data in a supervised way, we propose a multi-task learning framework by introducing an auxiliary task. In particular, we use leg side identification as an auxiliary task, which allows the use of more datasets, e.g., CHECK dataset. The work demonstrates that the utilization of additional data can improve the primary task of KL-grade classification for which only limited labelled data is available. This semi-supervised learning essentially helps to improve the feature learning ability of our framework, which leads to improved performance for KL-grade classification. We rigorously evaluated our proposed model on the two largest publicly available datasets for various aspects, i.e., overall performance, the effect of additional unlabelled samples and auxiliary tasks, robustness analysis, and ablation study. The proposed model achieved the accuracy, precision, recall, and F1 score of 75.53%, 74.1%, 78.51%, and 75.34%, respectively. Furthermore, the experimental results show that the suggested model not only achieves state-of-the-art performance on two publicly available datasets but also exhibits remarkable robustness.

Multitask deep learning on mammography to predict extensive intraductal component in invasive breast cancer.

To develop a multitask deep learning (DL) algorithm to automatically classify mammography imaging findings and predict the existence of extensive intraductal component (EIC) in invasive breast cancer. Mammograms with invasive breast cancers from 2010 to 2019 were downloaded for two radiologists performing image segmentation and imaging findings annotation. Images were randomly split into training, validation, and test datasets. A multitask approach was performed on the EfficientNet-B0 neural network mainly to predict EIC and classify imaging findings. Three more models were trained for comparison, including a single-task model (predicting EIC), a two-task model (predicting EIC and cell receptor status), and a three-task model (combining the abovementioned tasks). Additionally, these models were trained in a subgroup of invasive ductal carcinoma. The DeLong test was used to examine the difference in model performance. This study enrolled 1459 breast cancers on 3076 images. The EIC-positive rate was 29.0%. The three-task model was the best DL model with an area under the curve (AUC) of EIC prediction of 0.758 and 0.775 at the image and breast (patient) levels, respectively. Mass was the most accurately classified imaging finding (AUC = 0.915), followed by calcifications and mass with calcifications (AUC = 0.878 and 0.824, respectively). Cell receptor status prediction was less accurate (AUC = 0.625-0.653). The multitask approach improves the model training compared to the single-task model, but without significant effects. A mammography-based multitask DL model can perform simultaneous imaging finding classification and EIC prediction. The study results demonstrated the potential of deep learning to extract more information from mammography for clinical decision-making. • Extensive intraductal component (EIC) is an independent risk factor of local tumor recurrence after breast-conserving surgery. • A mammography-based deep learning model was trained to predict extensive intraductal component close to radiologists' reading. • The developed multitask deep learning model could perform simultaneous imaging finding classification and extensive intraductal component prediction.

Multitask Learning with Convolutional Neural Networks and Vision Transformers Can Improve Outcome Prediction for Head and Neck Cancer Patients.

Neural-network-based outcome predictions may enable further treatment personalization of patients with head and neck cancer. The development of neural networks can prove challenging when a limited number of cases is available. Therefore, we investigated whether multitask learning strategies, implemented through the simultaneous optimization of two distinct outcome objectives (multi-outcome) and combined with a tumor segmentation task, can lead to improved performance of convolutional neural networks (CNNs) and vision transformers (ViTs). Model training was conducted on two distinct multicenter datasets for the endpoints loco-regional control (LRC) and progression-free survival (PFS), respectively. The first dataset consisted of pre-treatment computed tomography (CT) imaging for 290 patients and the second dataset contained combined positron emission tomography (PET)/CT data of 224 patients. Discriminative performance was assessed by the concordance index (C-index). Risk stratification was evaluated using log-rank tests. Across both datasets, CNN and ViT model ensembles achieved similar results. Multitask approaches showed favorable performance in most investigations. Multi-outcome CNN models trained with segmentation loss were identified as the optimal strategy across cohorts. On the PET/CT dataset, an ensemble of multi-outcome CNNs trained with segmentation loss achieved the best discrimination (C-index: 0.29, 95% confidence interval (CI): 0.22-0.36) and successfully stratified patients into groups with low and high risk of disease progression (p=0.003). On the CT dataset, ensembles of multi-outcome CNNs and of single-outcome ViTs trained with segmentation loss performed best (C-index: 0.26 and 0.26, CI: 0.18-0.34 and 0.18-0.35, respectively), both with significant risk stratification for LRC in independent validation (p=0.002 and p=0.011). Further validation of the developed multitask-learning models is planned based on a prospective validation study, which has recently completed recruitment.

Comprehensive wheat lodging detection after initial lodging using UAV RGB images

Crop lodging in agricultural fields is one of the major factors that limit cereal crop yields. Wheat, the most popular cereal crop in most countries, is also affected by this phenomenon, which may result in a significant decrease in both yield and quality. Therefore, addressing wheat lodging is crucial for producers. This study aims to detect and identify wheat lodging through aerial images and classify its severity based on ratio, angle and location of the lodging. To achieve this goal, a multi-task approach was proposed involving three phases. First, automatic dataset generation methodology was conducted on orthomosaic imagery of three dates. Next, a comprehensive assessment of wheat lodging (ratio, angle and location) was performed, which has received little research attention. Third, applying and improving selected classification models for classifying image datasets was conducted. Combining convolutional neural networks and temporal sequences in a single model provided an opportunity to use spatiotemporal information extracted from the wheat image datasets. Time dependency and individual dates were both considered in the classification task. The limited number of data and imbalanced classes challenges, resulting from real field conditions data collection, were overcome by applying a new loss function to the classifier models. The overall accuracy of wheat lodging classification reached over 91% in these two states using the proposed approach. Based on this research, wheat lodging was detected more accurately by the proposed models despite the small and imbalanced dataset. The developed methodology paves the way for comprehensive and automatic wheat lodging detection, and the methodology can be adapted for similar crops that suffer lodging issues with suitable modifications.

Breaking the barriers of data scarcity in drug-target affinity prediction.

Accurate prediction of drug-target affinity (DTA) is of vital importance in early-stage drug discovery, facilitating the identification of drugs that can effectively interact with specific targets and regulate their activities. While wet experiments remain the most reliable method, they are time-consuming and resource-intensive, resulting in limited data availability that poses challenges for deep learning approaches. Existing methods have primarily focused on developing techniques based on the available DTA data, without adequately addressing the data scarcity issue. To overcome this challenge, we present the Semi-Supervised Multi-task training (SSM) framework for DTA prediction, which incorporates three simple yet highly effective strategies: (1) A multi-task training approach that combines DTA prediction with masked language modeling using paired drug-target data. (2) A semi-supervised training method that leverages large-scale unpaired molecules and proteins to enhance drug and target representations. This approach differs from previous methods that only employed molecules or proteins in pre-training. (3) The integration of a lightweight cross-attention module to improve the interaction between drugs and targets, further enhancing prediction accuracy. Through extensive experiments on benchmark datasets such as BindingDB, DAVIS and KIBA, we demonstrate the superior performance of our framework. Additionally, we conduct case studies on specific drug-target binding activities, virtual screening experiments, drug feature visualizations and real-world applications, all of which showcase the significant potential of our work. In conclusion, our proposed SSM-DTA framework addresses the data limitation challenge in DTA prediction and yields promising results, paving the way for more efficient and accurate drug discovery processes.

Range estimation in a weakly range-dependent waveguide in the South China Sea using single hydrophone and multi-task trained deep neural network

In this paper, an underwater source range estimation method using a multi-task trained deep neural network is proposed. Unlike existing machine learning frameworks, in this case, the physical quantities utilized for the training and testing are not identical. The training data parameter is the phase of the complex sound pressure, which is independent of the source and receiver depth and can be generated easily by an acoustic propagation model with only a few environmental parameters, decreasing computational costs significantly. The testing data parameter is the realistically-measured complex sound pressure. Instead of simply using the range as the only supervision signal, two other parameters — the phase mode and the bottom sound speed — are applied as auxiliary labels during the training process, embedding additional information about the sound field into the deep model. The projection between the sound field features and the range extracted from the training data by the deep model is applied to range estimation task on testing data. The effectiveness of the proposed approach is verified for a weakly range-dependent waveguide in the South China Sea by localizing several upsweep signals within a frequency band of 20–200 Hz. Testing results illustrate that the proposed method outperforms the classical matched impulse response approach and deep-learning-based single-label trained models, with the mean absolute percentage error (MAPE) decreasing by 24.63% and 59.46%, respectively. The mechanism of promotion is explained by comparing visualized hidden features of the neural network using t-distributed stochastic neighbor embedding (t-SNE) between the single-task and multi-task approach, showing that the latter helps to construct a more reasonable data manifold in high-dimensional space and thus ensures better performance of the deep model. We also fine-tune the network with a few realistic samples and further improve the performance by up to 25.34% against the baseline.

Multi-attention-based approach for deepfake face and expression swap detection and localization

Advancements in facial manipulation technology have resulted in highly realistic and indistinguishable face and expression swap videos. However, this has also raised concerns regarding the security risks associated with deepfakes. In the field of multimedia forensics, the detection and precise localization of image forgery has become essential tasks. Current deepfake detectors perform well with high-quality faces within specific datasets, but often struggle to maintain their performance when evaluated across different datasets. To this end, we propose an attention-based multi-task approach to improve feature maps for classification and localization tasks. The encoder and the attention-based decoder of our network generate localized maps that highlight regions with information about the type of manipulation. These localized features are shared with the classification network, improving its performance. Instead of using encoded spatial features, attention-based localized features from the decoder’s first layer are combined with frequency domain features to create a discriminative representation for deepfake detection. Through extensive experiments on face and expression swap datasets, we demonstrate that our method achieves competitive performance in comparison to state-of-the-art deepfake detection approaches in both in-dataset and cross-dataset scenarios. Code is available at https://github.com/saimawaseem/Multi-Attention-Based-Approach-for-Deepfake-Face-and-Expression-Swap-Detection-and-Localization.

Multi-tasking Federated Learning meets Blockchain to Foster Trust and Security in the Metaverse

The Metaverse is a term that refers to a shared virtual space where users can interact with each other and with a virtual environment in real-time. The Metaverse has gained a lot of attention in recent years owing to the accelerated advancements in virtual and augmented realty technologies, which made it possible to create more immersive and highly interactive experiences for users. However, as users engage with the Metaverse, they generate a lot of data, which may include their location, activity, and interactions with other users. This data can be used to train multiple machine learning (ML) models to personalize content and ads, but it can also be exploited for more nefarious uses. The problem of user data collection and the sharing of sensitive and private data in the Metaverse is a challenge. Federated learning (FL) is a distributed machine learning (ML) mechanism that can avoid problems related to data sharing by having multiple devices work together to create a ML model while maintaining their local training data private. The utilization of FL in the Metaverse involves numerous contributors, thereby increasing the susceptibility to malicious activity, including attacks. To address this issue, a promising solution is the integration of blockchain technology, which can provide a trusted solution to the metaverse that allows for secure model learning. Due to frequent communication between FL and blockchain, FL needs to be adapted to the limited communication bandwidth and limited resources of the device. In this paper, we propose a blockchain-based FL framework to enhance security and trustworthiness in the Metaverse. The proposed solution utilizes a multi-task FL approach that leverages blockchain sharding to enhance the system’s throughput while simultaneously reducing resource requirements. By breaking the blockchain network into smaller shards, the processing power required for each shard is reduced, leading to an increase in overall throughput. The multi-task FL approach allows for the simultaneous training of multiple ML models, reducing the time and resources required to train each model individually. We formulate the model learning problem into a blockchain sharding problem and propose a bipartite matching solution for shard creation. Furthermore, we propose a scheduling approach to distribute the bandwidth between reliable devices, hence minimizing communication across FL devices and giving devices with a reliable behavior and informative dataset priority. Numerical results demonstrated that our framework has better performance than referenced solutions on the selected indicators.

Deep learning-driven diagnosis: A multi-task approach for segmenting stroke and Bell's palsy

Strong efforts have been undertaken to enhance the diagnosis and identification of diseases that cause facial paralysis, such as Bell's palsy and stroke, because of their detrimental social effects. Stroke is one of the most serious and potentially fatal conditions among the major cardiovascular disorders. We are introducing a deep-learning-based method for early diagnosis of facial paralysis diseases such as stroke and Bell's palsy. Recognizing the costs associated with traditional diagnostic techniques like magnetic resonance tomography (MRI) and computed tomography (CT) scan images, our model employs a multi-task network, integrating face parsing, facial asymmetry parsing, and category enhancement. Spatial inconsistencies are addressed via a depth-map estimation module that leverages an instance-specific kernel approach. To clarify the boundaries of facial components, we use category edge detection with a foreground attention module, generating generic geometric structures and detailed semantic cues. Our model is trained on two datasets, comprising individuals with regular smiles and those with one-sided facial weakness. This cost-effective, easily accessible solution can streamline the diagnostic process, minimizing data gaps, and reducing needless rescreening and intervention costs.

DeepTraSynergy: drug combinations using multimodal deep learning with transformers.

Screening bioactive compounds in cancer cell lines receive more attention. Multidisciplinary drugs or drug combinations have a more effective role in treatments and selectively inhibit the growth of cancer cells. Hence, we propose a new deep learning-based approach for drug combination synergy prediction called DeepTraSynergy. Our proposed approach utilizes multimodal input including drug-target interaction, protein-protein interaction, and cell-target interaction to predict drug combination synergy. To learn the feature representation of drugs, we have utilized transformers. It is worth noting that our approach is a multitask approach that predicts three outputs including the drug-target interaction, its toxic effect, and drug combination synergy. In our approach, drug combination synergy is the main task and the two other ones are the auxiliary tasks that help the approach to learn a better model. In the proposed approach three loss functions are defined: synergy loss, toxic loss, and drug-protein interaction loss. The last two loss functions are designed as auxiliary losses to help learn a better solution. DeepTraSynergy outperforms the classic and state-of-the-art models in predicting synergistic drug combinations on the two latest drug combination datasets. The DeepTraSynergy algorithm achieves accuracy values of 0.7715 and 0.8052 (an improvement over other approaches) on the DrugCombDB and Oncology-Screen datasets, respectively. Also, we evaluate the contribution of each component of DeepTraSynergy to show its effectiveness in the proposed method. The introduction of the relation between proteins (PPI networks) and drug-protein interaction significantly improves the prediction of synergistic drug combinations. The source code and data are available at https://github.com/fatemeh-rafiei/DeepTraSynergy.

Hypernetworks for Zero-Shot Transfer in Reinforcement Learning

In this paper, hypernetworks are trained to generate behaviors across a range of unseen task conditions, via a novel TD-based training objective and data from a set of near-optimal RL solutions for training tasks. This work relates to meta RL, contextual RL, and transfer learning, with a particular focus on zero-shot performance at test time, enabled by knowledge of the task parameters (also known as context). Our technical approach is based upon viewing each RL algorithm as a mapping from the MDP specifics to the near-optimal value function and policy and seek to approximate it with a hypernetwork that can generate near-optimal value functions and policies, given the parameters of the MDP. We show that, under certain conditions, this mapping can be considered as a supervised learning problem. We empirically evaluate the effectiveness of our method for zero-shot transfer to new reward and transition dynamics on a series of continuous control tasks from DeepMind Control Suite. Our method demonstrates significant improvements over baselines from multitask and meta RL approaches.

Named Entity Recognition in Russian Using Multi-Task LSTM-CRF

Named entity recognition (NER) is aimed at obtaining the important information from the unstructured data presented in the form of natural language texts. In this work, we investigate the efficiency of modern multi-task NER approaches on Russian language corpora by employing several different NER datasets and a dataset of part-of-speech (POS) tags. We apply a state-of-the-art neural architecture based on bidirectional LSTMs and conditional random fields. Convolutional neural networks were utilized to learn character-level features. We carry out extensive experimental evaluation over three standard datasets of news articles written in Russian. The proposed multi-task model achieve state-of-the-art results with an F1 score of 88.04% on Gareev’s dataset and an F1 score of 99.49% on Person-1000 dataset.

Automated LVO detection and collateral scoring on CTA using a 3D self-configuring object detection network: a multi-center study

The use of deep learning (DL) techniques for automated diagnosis of large vessel occlusion (LVO) and collateral scoring on computed tomography angiography (CTA) is gaining attention. In this study, a state-of-the-art self-configuring object detection network called nnDetection was used to detect LVO and assess collateralization on CTA scans using a multi-task 3D object detection approach. The model was trained on single-phase CTA scans of 2425 patients at five centers, and its performance was evaluated on an external test set of 345 patients from another center. Ground-truth labels for the presence of LVO and collateral scores were provided by three radiologists. The nnDetection model achieved a diagnostic accuracy of 98.26% (95% CI 96.25–99.36%) in identifying LVO, correctly classifying 339 out of 345 CTA scans in the external test set. The DL-based collateral scores had a kappa of 0.80, indicating good agreement with the consensus of the radiologists. These results demonstrate that the self-configuring 3D nnDetection model can accurately detect LVO on single-phase CTA scans and provide semi-quantitative collateral scores, offering a comprehensive approach for automated stroke diagnostics in patients with LVO.

Multitask Approach to Localize Rhizobial Type Three Secretion System Effector Proteins Inside Eukaryotic Cells.

Rhizobia can establish mutually beneficial interactions with legume plants by colonizing their roots to induce the formation of a specialized structure known as a nodule, inside of which the bacteria are able to fix atmospheric nitrogen. It is well established that the compatibility of such interactions is mainly determined by the bacterial recognition of flavonoids secreted by the plants, which in response to these flavonoids trigger the synthesis of the bacterial Nod factors that drive the nodulation process. Additionally, other bacterial signals are involved in the recognition and the efficiency of this interaction, such as extracellular polysaccharides or some secreted proteins. Some rhizobial strains inject proteins through the type III secretion system to the cytosol of legume root cells during the nodulation process. Such proteins, called type III-secreted effectors (T3E), exert their function in the host cell and are involved, among other tasks, in the attenuation of host defense responses to facilitate the infection, contributing to the specificity of the process. One of the main challenges of studying rhizobial T3E is the inherent difficulty in localizing them in vivo in the different subcellular compartments within their host cells, since in addition to their low concentration under physiological conditions, it is not always known when or where they are being produced and secreted. In this paper, we use a well-known rhizobial T3E, named NopL, to illustrate by a multitask approach where it localizes in heterologous hosts models, such as tobacco plant leaf cells, and also for the first time in transfected and/or Salmonella-infected animal cells. The consistency of our results serves as an example to study the location inside eukaryotic cells of effectors in distinct hosts with different handling techniques that can be used in almost every research laboratory.

A multi-task learning-based optimization approach for finding diverse sets of microstructures with desired properties

Optimization along the chain processing-structure-properties-performance is one of the core objectives in data-driven materials science. In this sense, processes are supposed to manufacture workpieces with targeted material microstructures. These microstructures are defined by the material properties of interest and identifying them is a question of materials design. In the present paper, we addresse this issue and introduce a generic multi-task learning-based optimization approach. The approach enables the identification of sets of highly diverse microstructures for given desired properties and corresponding tolerances. Basically, the approach consists of an optimization algorithm that interacts with a machine learning model that combines multi-task learning with siamese neural networks. The resulting model (1) relates microstructures and properties, (2) estimates the likelihood of a microstructure of being producible, and (3) performs a distance preserving microstructure feature extraction in order to generate a lower dimensional latent feature space to enable efficient optimization. The proposed approach is applied on a crystallographic texture optimization problem for rolled steel sheets given desired properties.

Deep reinforcement learning based adaptive threshold multi-tasks offloading approach in MEC

Offloading tasks and data is an important use case for Mobile Edge Computing (MEC) that benefits a wide variety of mobile applications on different platforms, including autonomous vehicles and smartphones. A moving vehicle may pass through multiple Roadside Units (RSUs) and the MEC servers located within them as it travels. On-board applications generate structured tasks and can offload them to observable servers. However, since the mobile node cannot determine the information of the global server, choosing the best server becomes a key issue. In this paper, we address the problem of the offloading threshold in traditional Optimal Stopping Theory (OST) based decision models that cannot adapt to the environment. We use Deep Q-Network (DQN) to provide the model with the optimal offloading threshold for adaptive environments and propose a DQN-based Adaptive Threshold Multi-task Offloading approach (DQN-ATMOO). This method uses the principle of OST to solve the offloading order decision problem, divides structured tasks into sub-tasks, and executes offloading in sequence, with the goal of maximizing the probability of offloading to the best server and minimizing the total offloading delay. We experimentally compare and evaluate the proposed model using simulated and real datasets. The results show that the proposed model can be effectively implemented on mobile nodes and significantly reduces the total expected processing time.

Joint segmentation and classification of skin lesions via a multi-task learning convolutional neural network

Skin lesion segmentation and classification are two crucial and correlated tasks in computer-aided diagnosis of skin diseases. Jointly performing these two tasks can exploit their correlations to obtain performance gains, but it remains a challenging topic. In this paper, we propose an end-to-end multi-task learning convolutional neural network (MTL-CNN) for joint skin lesion segmentation and classification, and additionally introduce edge prediction as an auxiliary task. Overall, MTL-CNN includes a shared encoder, two parallel decoders for generating edge and segmentation masks, and a classification subnet. First, the shared encoder is used to extract features for three tasks (i.e., edge prediction, segmentation, and classification). Then, we propose two kinds of simple but efficient modules to exploit the benefits among these three tasks. Specifically, we design multiple edge information enhancement (EIE) modules between the encoder and the segmentation decoder, aiming at introducing the edge information from the edge decoder as strong cues to enhance the edge parts of the segmentation features. These enhanced segmentation features are sent to the segmentation decoder for better segmentation. Besides, we design multiple lesion area extraction (LAE) modules between the encoder and the classification subnet, which aim to utilize the segmentation results to filter out the distractions on the classification features. These filtered classification features are input to the classification subnet and progressively fused in a bottom-up manner for classification. A three-phase training strategy is employed to train MTL-CNN. Extensive experiments on three datasets demonstrate the superiority of MTL-CNN over state-of-the-art segmentation, classification, and other multi-task approaches.