ABSTRACT Background Modern critical systems often require to take decisions and classify data and scenarios autonomously without having detrimental effects on people, infrastructures or the environment, ensuring desired dependability attributes. Researchers typically strive to craft classifiers with perfect accuracy, which should be always correct and as such never threaten the encompassing system. Unfortunately, this is a very unrealistic goal, as classification tasks are typically complex and may encounter a wide variety of unexpected operating conditions and unknown inputs. Methods Classifiers should be considered as building blocks that interact with other components that help rejecting those predictions that are suspected to be misclassifications, triggering system‐level mitigation strategies instead. Fail‐Controlled Classifiers (FCCs) are software components that can either correctly classify, misclassify, or reject outputs: ideally, they would reject all and only outputs that correspond to misclassifications. Nine different FCCs are presented: Self‐Checking Classifiers (SCC), Watchdog Timers (WT), Input Processor (IP), Output processor (OP), Safety Wrapper (SW), Recovery Blocks (RB), weighted and non‐weighted Voting (VT, WVT) and Stacking (STK). Results These 9 FCCs are instantiated in experiments with tabular and image classifiers, showing their potential in rejecting most misclassifications and paving the ways for trustworthy decisions to be deployed in critical systems. If the system can tolerate more omissions, the IP FCC is a good choice. On the other hand, if achieving the highest accuracy is the priority, RB FCC performs better. Conclusions Findings show that FCCs do not primarily aim at improving correct classifications, but allow for transforming many misclassifications into rejections, which may be easily handled by the encompassing system and paving the way for trustworthy decisions to be deployed in critical systems.

Federated Learning (FL) stands out in the realm of collaborative learning that ensures data privacy for each client involved in the federation. Despite showing astounding potential, it inherently brings challenges that are often difficult to address and make it hardly applicable in industrial scenarios. First, all clients must agree on the same algorithm to be trained locally to enable global averaging, limiting autonomy. Second, clients are required to disclose insights of their models (i.e., weights or gradients) to create the global model upon averaging and aggregating techniques, threatening privacy. Third, applications of FL in critical systems where each component has to be trusted are not solid enough at best. This paper designs Trustworthy, Black-Box FL (TBB-FL) software architectures that allow clients to locally train any algorithm they want to solve a specific task. Only executables of local models are sent to the server and herein treated as black-boxes to create the global model as an adjudication of clients' opinions. Moreover, local and global models are self-checking, software components that quantify confidence in a prediction to suspect prediction errors, ultimately rejecting outputs that cannot be trusted. We validate this approach through classification experiments on both image and tabular datasets using the Flower framework, comparing TBB-FL against traditional FL and against individual local models. TBB-FL heavily reduces misclassifications compared to traditional FL, with minimal accuracy drop, and has better classification performance than local models alone.

Abstract Text, as a vital tool for communication, is playing an imperative role in modern society. Precise high-level text translation systems are essential requirements in a wide range of real-world applications, such as robot navigation, industrial automation, image search, and instant translation. Regardless of improved research, a series of grand challenges may still become upon when translating text automatically in the real-world from open scene images. The difficulties mainly stem from multiplicity and inconsistency of text in open scenes, complication and obstruction of backgrounds, and deficient imaging conditions in uncontrolled circumstances for open scene images. The existing deep learning-based text translation systems do not eliminate the text for translation, and these applications just replace text on the reconstructed scene. To address the abovementioned shortcomings, this study proposed a novel approach for open scene text translation. Our system consists of five modules including scene text detection, text recognition, text elimination, text translation, and text insertion along with scene reconstruction. The novelty presented by our model lies in the idea of first eliminating the text from the open scene for accurate translation and then reconstructs the translated text on the image for its proper alignment. We specifically modified the existing generative adversarial network (GAN) architecture for improved performance of text elimination by introducing a novel strategy of text and scene concatenation to reduce the overall loss function. For this purpose, we created a synthetic dataset to train our GAN for text elimination module. Experiments on various standard text translation systems demonstrate that our integrated system is able to outperform state-of-the-art approaches in terms of result quality. We have achieved 90.87% of precision, 83.66% of recall, 87.116% of F1-score, and reduced both losses ( $$l_1$$ l 1 and $$l_2$$ l 2 ) up to 50% which is remarkable upon state-of-the-art translation systems.

Abstract Water is essential for human survival. Humans can live without food for a few days but without water, a person can barely survive for 3–5 days. Various parts of the world, particularly under-developed countries, have regions where clean water is scarce, and humans living in such conditions have no access to clean water. Our solution provides information on whether water is contaminated or not. Moreover, it overcomes the delay time in getting the result of water contamination using traditional methods of up to 5–6 hrs. Our proposed method detects the colonies of the bacteria that are taken from the water sample (after gram staining) and then classifies the type of bacteria to whom it belongs and how much quantity of each bacterium causes infection to the human body. Bacteria detection is performed by a novel deep learning-based model with user-specified parameters. To improve our ability to detect dangerous bacteria including E. coli, yeast, and particles, we perform tests using datasets from a variety of researchers. On the test benchmark, the fine-tuned proposed model achieves 84.56% accuracy and provides the level of contamination in water.

Domain experts are desperately looking to solve decision-making problems by designing and training Machine Learning algorithms that can perform classification with the highest possible accuracy. No matter how hard they try, classifiers will always be prone to misclassifications due to a variety of reasons that make the decision boundary unclear. This complicates the integration of classifiers into critical systems, where misclassifications could directly impact people, infrastructures, or the environment. The paper proposes to consider a classifier as a structural part of the system instead of an individual component to be tested in isolation and included in the system afterward. This allows for omitting those predictions that are suspected to be misclassifications, triggering system-level mitigation strategies. The resulting fail-controlled classifiers (FCCs) are software components that can correctly classify, misclassify, or omit outputs: ideally, they would omit all and only outputs that correspond to misclassifications. After presenting the theoretical foundations of FCCs, the paper proposes metrics to quantify their performance, 5 software architectures for FCCs, and an experimental analysis involving tabular data and image classifiers. Overall, this paper advocates the need for a system and software design in which ML classifiers are not separate components, but should rather be considered building blocks that interact with other components for improved performance.

Abstract Employing cognitive robots, capable of throwing and catching, is a strategy aimed at expediting the logistics process within Industry 4.0’s smart manufacturing plants, specifically for the transportation of small-sized manufacturing parts. Since the flight of mechanically thrown objects is inherently unpredictable, it is crucial for the catching robot to observe the initial trajectory with utmost precision and intelligently forecast the final catching position to ensure accurate real-time grasping. This study utilizes multi-camera tracking to monitor mechanically thrown objects. It involves the creation of a 3D simulation that facilitates controlled mechanical throwing of objects within the internal logistics environment of Industry 4.0. The developed simulation empowers users to define the attributes of the thrown object and capture its trajectory using a simulated pinhole camera, which can be positioned at any desired location and orientation within the in-plant logistics environment of flexible manufacturing systems. The simulation facilitated ample experimentation to be conducted for determining the optimal camera positions for accurately observing the 3D interception positions of a flying object based on its apparent size on the camera’s sensor plane. Subsequently, a variety of calibrated multi-camera setups were experimented while placing cameras at identified optimal positions. Based on the obtained results, the most effective multi-camera configuration setup is derived. Finally, a training dataset is prepared for 3000 simulated throwing experiments where the initial part of the trajectory consists of observed interception positions, through derived best multi-camera setup, and the final part consists of actual positions. The encoder–decoder Bi-LSTM deep neural network is proposed and trained on this dataset. The trained model outperformed the current state-of-the-art by accurately predicting the final 3D catching point, achieving a mean average error of 5 mm and a root-mean-square error of 7 mm in 200 real-world test experiments.