Convolution neural network (CNN) is one of the most popular machine learning techniques that is being used in many applications like image classification, image analysis, textile archives, object recognition, and many more. In the textile industry, the classification of defective and nondefective fabric is an essential and necessary step to control the quality of fabric. Traditionally, a user physically inspects and classifies the fabric, which is an ineffective and tedious activity. Therefore, it is desirable to have an automated system for detecting defects in the fabric. To address these issues, this research proposes a solution for classifying defective and nondefective fabric using deep learning-based framework. Therefore, in this research, an image processing technique with CNN-based GoogleNet is presented to classify defective and nondefective fabric. To achieve the purpose, the system is trained using different kinds of fabric defects. The performance of the proposed approach was evaluated on the textile texture TILDA dataset, and achieved a classification accuracy of 94.46%. The classification results show that the proposed approach for classifying defective and nondefective fabric is better as compared to other state-of-the-art approaches such as Bayesian, BPNN, and SVM.

Advancement in technology has led to an increase in data. Consequently, techniques such as deep learning and artificial intelligence which are used in deciphering data are increasingly becoming popular. Further, advancement in technology does increase user expectations on devices, including consumer interfaces such as mobile apps, virtual environments, or popular software systems. As a result, power from the battery is consumed fast as it is used in providing high definition display as well as in charging the sensors of the devices. Low latency requires more power consumption in certain conditions. Cloud computing improves the computational difficulties of smart devices with offloading. By optimizing the device’s parameters to make it easier to find optimal decisions for offloading tasks, using a metaheuristic algorithm to transfer the data or offload the task, cloud computing makes it easier. In cloud servers, we offload the tasks and limit their resources by simulating them in a virtual environment. Then we check resource parameters and compare them using metaheuristic algorithms. When comparing the default algorithm FCFS to ACO or PSO, we find that PSO has less battery or makespan time compared to FCFS or ACO. The energy consumption of devices is reduced if their resources are offloaded, so we compare the results of metaheuristic algorithms to find less battery usage or makespan time, resulting in the PSO increasing battery life or making the system more efficient.

Deep-learning research on flood inundation mapping is driven largely by architectural innovation and single-run reporting. We ask whether the segmentation architecture or the input representation contributes more to accuracy in flood-event surface-water mapping. We introduce GeoHybridNet, a hybrid CNN–Transformer network, and compare it on the hand-labeled Sen1Floods11 benchmark with a U-Net, an Attention U-Net, a parameter-matched wide U-Net, and SegFormer under matched data and training. The four convolutional and hybrid models differ by at most 1.4 IoU points, and none are significantly different under event-level tests; a compact SegFormer trained from scratch is about 3.5 points lower. In factorial designs crossing three architectures with up to five input representations (five seeds each), input representation explains about 95% of the variance in water IoU and architecture about 1%. The dominant input factor is radiometric normalization: raw bands standardized per chip reach only 60% IoU and adding NDWI restores 76%, whereas the same bands standardized with training-set statistics reach 77.5% without any index. The pattern holds when permanent water is excluded. On three held-out flood events, accuracy drops by 24–26 IoU points for every model. On this benchmark, input preparation and honest evaluation matter more than architecture.

Background/Objectives: Machine-learning studies of autism spectrum disorder from resting-state functional magnetic resonance imaging routinely report near-perfect accuracy, yet the models rarely transfer to an unseen scanner. We audited a pipeline of the kind this literature builds, which reported an accuracy of 0.999, and traced that number to five data-leakage pathways. Methods: Under a nested leave-one-site-out protocol on 871 participants from the Autism Brain Imaging Data Exchange, in which the held-out site influenced nothing but model selection, the threshold, the combiner, or the site correction, we closed every pathway, priced each by reopening it, and validated on a locked external cohort. Results: With all five pathways closed, the graph attention network reached an area under the curve of 0.648 and did not outperform a support vector machine on the same connectivity (0.679; p = 0.009). Decision-level fusion added nothing detectable (0.682, p = 0.72). A network identified its training participants better than it detected autism (0.710 versus 0.610); both questionnaire benchmarks proved label-tautological, and performance held on the external cohort. Conclusions: Reported gains in this literature should be read as claims about evaluation protocols until shown otherwise on an untouched acquisition site; all predictions, models, and code are released.