Filtern
Erscheinungsjahr
- 2022 (85) (entfernen)
Dokumenttyp
Schlagworte
- CDK (2)
- Journalismus (2)
- Marketing (2)
- Air handling unit (1)
- Automatisierung (1)
- BIM (1)
- Befragung (1)
- Berufsfeldstudie (1)
- Biomechanics (1)
- Brand theory (1)
The development of deep learning-based optical chemical structure recognition (OCSR) systems has led to a need for datasets of chemical structure depictions. The diversity of the features in the training data is an important factor for the generation of deep learning systems that generalise well and are not overfit to a specific type of input. In the case of chemical structure depictions, these features are defined by the depiction parameters such as bond length, line thickness, label font style and many others. Here we present RanDepict, a toolkit for the creation of diverse sets of chemical structure depictions. The diversity of the image features is generated by making use of all available depiction parameters in the depiction functionalities of the CDK, RDKit, and Indigo. Furthermore, there is the option to enhance and augment the image with features such as curved arrows, chemical labels around the structure, or other kinds of distortions. Using depiction feature fingerprints, RanDepict ensures diversely picked image features. Here, the depiction and augmentation features are summarised in binary vectors and the MaxMin algorithm is used to pick diverse samples out of all valid options. By making all resources described herein publicly available, we hope to contribute to the development of deep learning-based OCSR systems.
The development of deep learning-based optical chemical structure recognition (OCSR) systems has led to a need for datasets of chemical structure depictions. The diversity of the features in the training data is an important factor for the generation of deep learning systems that generalise well and are not overfit to a specific type of input. In the case of chemical structure depictions, these features are defined by the depiction parameters such as bond length, line thickness, label font style and many others. Here we present RanDepict, a toolkit for the creation of diverse sets of chemical structure depictions. The diversity of the image features is generated by making use of all available depiction parameters in the depiction functionalities of the CDK, RDKit, and Indigo. Furthermore, there is the option to enhance and augment the image with features such as curved arrows, chemical labels around the structure, or other kinds of distortions. Using depiction feature fingerprints, RanDepict ensures diversely picked image features. Here, the depiction and augmentation features are summarised in binary vectors and the MaxMin algorithm is used to pick diverse samples out of all valid options. By making all resources described herein publicly available, we hope to contribute to the development of deep learning-based OCSR systems.
The use of molecular string representations for deep learning in chemistry has been steadily increasing in recent years. The complexity of existing string representations, and the difficulty in creating meaningful tokens from them, lead to the development of new string representations for chemical structures. In this study, the translation of chemical structure depictions in the form of bitmap images to corresponding molecular string representations was examined. An analysis of the recently developed DeepSMILES and SELFIES representations in comparison with the most commonly used SMILES representation is presented where the ability to translate image features into string representations with transformer models was specifically tested. The SMILES representation exhibits the best overall performance whereas SELFIES guarantee valid chemical structures. DeepSMILES perform in between SMILES and SELFIES, InChIs are not appropriate for the learning task. All investigations were performed using publicly available datasets and the code used to train and evaluate the models has been made available to the public.
We investigated the formation of Artemia franciscana swarms of freshly hatched instar I nauplii larvae. Nauplii were released into light gradients but then interrupted by light-direction changes, small obstacles, or long barriers. All experiments were carried out horizontally. Each experiment used independent replicates. Freshly produced Artemia broods were harvested from independent incubators thus providing true replicate cohorts of Artemia subjected as replicates to the experimental treatments.
We discovered that Artemia nauplii swarms can: 1. repeatedly react to non-obstructed light gradients that undergo repeated direction-changes and do so in a consistent way, 2. find their way to a light source within maze-like arrangements made from small transparent obstacles, 3. move as a swarm around extended transparent barriers, following a light gradient. This paper focuses on the recognition of whole-swarm behaviors, the description thereof and the recognition of differences in whole-swarm movements comparing non-obstructed swarming with swarms encountering obstacles. Investigations of the within-swarm behaviors of individual Artemia nauplii and their interactions with neighboring nauplii are in progress, e.g. in order to discover the underlying swarming algorithms and differences
thereof comparing non-obstructed vs. obstructed pathways.
Different charge treatment approaches are examined for cyclotide-induced plasma membrane disruption by lipid extraction studied with dissipative particle dynamics. A pure Coulomb approach with truncated forces tuned to avoid individual strong ion pairing still reveals hidden statistical pairing effects that may lead to artificial membrane stabilization or distortion of cyclotide activity depending on the cyclotide’s charge state. While qualitative behavior is not affected in an apparent manner, more sensitive quantitative evaluations can be systematically biased. The findings suggest a charge smearing of point charges by an adequate charge distribution. For large mesoscopic simulation boxes, approximations for the Ewald sum to account for mirror charges due to periodic boundary conditions are of negligible influence.
Nowadays, robots are found in a growing number of areas where they collaborate closely with humans. Enabled by lightweight materials and safety sensors, these cobots are gaining increasing popularity in domestic care, where they support people with physical impairments in their everyday lives. However, when cobots perform actions autonomously, it remains challenging for human collaborators to understand and predict their behavior, which is crucial for achieving trust and user acceptance. One significant aspect of predicting cobot behavior is understanding their perception and comprehending how they “see” the world. To tackle this challenge, we compared three different visualization techniques for Spatial Augmented Reality. All of these communicate cobot perception by visually indicating which objects in the cobot’s surrounding have been identified by their sensors. We compared the well-established visualizations Wedge and Halo against our proposed visualization Line in a remote user experiment with participants suffering from physical impairments. In a second remote experiment, we validated these findings with a broader non-specific user base. Our findings show that Line, a lower complexity visualization, results in significantly faster reaction times compared to Halo, and lower task load compared to both Wedge and Halo. Overall, users prefer Line as a more straightforward visualization. In Spatial Augmented Reality, with its known disadvantage of limited projection area size, established off-screen visualizations are not effective in communicating cobot perception and Line presents an easy-to-understand alternative.
Einleitung und Fragestellung
Zahlreiche empirische Befunden sprechen für die positiven Effekte authentischer Führung. Wir untersuchen ihre Antezedenzien.
Theoretischer Hintergrund
Authentische Führung meint Handeln im Einklang mit moralischen Werten. Aus sozialkognitiver Perspektive bezeichnet moralische Identität eine komplexe Wissensstruktur aus moralischen Werten, Zielen und Verhaltensmustern, welche durch Lebenserfahrungen erworben werden. Darin sehen wir eine Basis für authentische
Führung (H1). Sich trotz sozialer Opposition für moralische Prinzipien einzusetzen, ist bezeichnend für Mut. Dieser zeigt sich in selbstkongruentem Verhalten trotz negativer
Konsequenzen. Dem Identitätsprozessmodell folgend, wird Mut notwendig, wenn Identiätsspannungen Inkongruenz zwischen Selbstkonzept und sozialer Identität hervorrufen. Darin sehen wir ein Aktivierungspotenzial für authentische Führung (H2).
Methode
Wir befragten N = 70 Führungsdyaden eines Industriekonzerns. Mut (WSCS; Howard et al., 2016) und moralische Identität(MIS; Aquino & Reed, 2002) wurden als Selbsteinschätzung der Führungskräfte erhoben (Altersdurchschnitt: 46 Jahre, 59% ♂). Authentische Führung (ALQ, Walumbwa et al., 2008) erfassten wir als Fremdeinschätzung durch Mitarbeitende (Altersdurchschnitt: 37, 47% ♂).
Ergebnisse
Moralische Identität und tatsächliches Verhalten müssen scheinbar nicht notwendigerweise übereinstimmen; etwa wenn hohe Kosten für moralisches Verhalten erwartbar sind. Hier setzt sozialer Mut im Arbeitskontext an. Entsprechend
wird eine mutig agierende Führungskraft als authentisch wahrgenommen, vor allem, wenn dieses Verhalten mögliche negative soziale Konsequenzen beinhaltet.
Diskussion
Mutiges Handeln wird durch Persönlichkeit, Selbstwirksamkeit und aktuelle Emotionen geleitet und kann etwa in der Führungskräfteentwicklung gelernt werden.
Hier bieten sich narrative Formate an, die die Selbstreflexion fördern. Auch bzgl. der Entwicklung authentischer Führung verweisen erste Befunde auf die Bedeutung der persönlichen Reflexion, z.B. über die eigene Lebensgeschichte.