Search

Plankton communities form the basis of aquatic ecosystems and elucidating their role in increasingly important environmental issues is a persistent research question. Recent technological advances in automated microscopic imaging, together with cloud platforms for high-performance computing, have created possibilities for collecting and processing detailed high-frequency data on planktonic communities, opening new horizons for testing core hypotheses in aquatic ecosystems. Analyzing continuous streams of big data calls for development and deployment of novel computer vision and machine learning systems. The implementation of these analysis systems is not always straightforward with regards to operationality, and issues regarding data flows, computing and data treatment need to be considered. We created a data pipeline for automated near-real-time classification of phytoplankton during remote deployment of imaging flow cytometer (Imaging FlowCytobot, IFCB). Convolutional neural network (CNN) is used to classify continuous imaging data with probability thresholds used to filter out images not belonging to our existing classes. The automated data flow and classification system were used to monitor dominating species of filamentous cyanobacteria on the coast of Finland during summer 2021. We demonstrate that good phytoplankton recognition can be achieved with transfer learning utilizing a relatively shallow, publicly available, pre-trained CNN model and fine-tuning it with community-specific phytoplankton images (overall F1-score of 0.95 for test set of our labeled image data complemented with a 50% unclassifiable image portion). This enables both fast training and low computing resource requirements for model deployment making it easy to modify and applicable in wide range of situations. The system performed well when used to classify a natural phytoplankton community over different seasons (overall F1-score 0.82 for our evaluation data set). Furthermore, we address the key challenges of image classification for varying planktonic communities and analyze the practical implications of confused classes. We published our labeled image data set of Baltic Sea phytoplankton community for the training of image recognition models (~63000 images in 50 classes) to accelerate implementation of imaging systems for other brackish and freshwater communities. Our evaluation data set, 59 fully annotated samples of natural communities throughout an annual cycle, is also available for model testing purposes (~150000 images).

This paper discusses contemporary advancements in the affective sciences (described together as skeptical theories) that can inform the music-emotion literature. Key concepts in these theories are outlined, highlighting their points of agreement and disagreement. This summary shows the importance of appraisal within the emotion process, provides a greater emphasis upon goal-directed accounts of (emotion) behavior, and a need to move away from discrete emotion “folk” concepts and toward the study of an emotional episode and its components. Consequently, three contemporary music emotion theories (BRECVEMA, Multifactorial Process Approach, and a Constructionist Account) are examined through a skeptical lens. This critique highlights the over-reliance upon categorization and a lack of acknowledgment of appraisal processes, specifically goal-directed appraisal, in examining how individual experiences of music emerge in different contexts. Based on this critique of current music-emotion models, we present our skeptically informed CODA model - Constructivistly-Organised Dimensional-Appraisal model. This model addresses skeptical limitations of existing theories, reinstates the role of goal-directed appraisal as central to what makes music relevant and meaningful to an individual in different contexts and brings together different theoretical frameworks into a single model. From the development of the CODA model, several hypotheses are proposed and applied to musical contexts. These hypotheses address theoretical issues such as acknowledging individual and contextual differences in emotional intensity and valence, as well as differentiating between induced and perceived emotions, and utilitarian and aesthetic emotions. We conclude with a sections of recommendations for future research. Altogether, this theoretical critique and proposed model points toward a positive future direction for music-emotion science. One where researchers can take forward testable predictions about what makes music relevant and meaningful to an individual.

Research on musical chills has linked the response to multiple musical features; however, there exists no study that has attempted to manipulate musical stimuli to enable causal inferences, meaning current understanding is based mainly on correlational evidence. In the current study, participants who regularly experience chills (N = 24) listened to an original and manipulated version of three pieces reported to elicit chills in a previous survey. Predefined chills sections were removed to create manipulated conditions. The effects of these manipulations on the chills response were assessed through continuous self-reports, and skin conductance measurements. Results show that chills were significantly less frequent following stimulus manipulation across all three pieces. Continuous measurements of chills intensity were significantly higher in the chills sections compared with control sections in the pieces; similar patterns were found for phasic skin conductance, although some differences emerged. Continuous measurements also correlated with psychoacoustic features such as loudness, brightness and roughness in two of the three pieces. Findings are discussed in terms of understanding structural and acoustic features and chills experiences within their local music contexts, the necessity of experimental approaches to musical chills, and the possibility of different features activating different underlying mechanisms.

Basic Emotion theory has had a tremendous influence on the affective sciences, including music psychology, where most researchers have assumed that music expressivity is constrained to a limited set of basic emotions. Several scholars suggested that these constrains to musical expressivity are explained by the existence of a shared acoustic code to the expression of emotions in music and speech prosody. In this article we advocate for a shift from this focus on basic emotions to a constructionist account. This approach proposes that the phenomenon of perception of emotions in music arises from the interaction of music’s ability to express core affects and the influence of top-down and contextual information in the listener’s mind. We start by reviewing the problems with the concept of Basic Emotions, and the inconsistent evidence that supports it. We also demonstrate how decades of developmental and cross-cultural research on music and emotional speech have failed to produce convincing findings to conclude that music expressivity is built upon a set of biologically pre-determined basic emotions. We then examine the cue-emotion consistencies between music and speech, and show how they support a parsimonious explanation, where musical expressivity is grounded on two dimensions of core affect (arousal and valence). Next, we explain how the fact that listeners reliably identify basic emotions in music does not arise from the existence of categorical boundaries in the stimuli, but from processes that facilitate categorical perception, such as using stereotyped stimuli and close-ended response formats, psychological processes of construction of mental prototypes, and contextual information. Finally, we outline our proposal of a constructionist account of perception of emotions in music, and spell out the ways in which this approach is able to make solve past conflicting findings. We conclude by providing explicit pointers about the methodological choices that will be vital to move beyond the popular Basic Emotion paradigm and start untangling the emergence of emotional experiences with music in the actual contexts in which they occur.

The paradox of enjoying listening to music that evokes sadness is yet to be fully understood. Unlike prior studies that have explored potential explanations related to lyrics, memories, and mood regulation, we investigated the types of emotions induced by unfamiliar, instrumental sad music, and whether these responses are consistently associated with certain individual difference variables. One hundred and two participants were drawn from a representative sample to minimise self-selection bias. The results suggest that the emotional responses induced by unfamiliar sad music could be characterized in terms of three underlying factors: Relaxing sadness, Moving sadness, and Nervous sadness. Relaxing sadness was characterized by felt and perceived peacefulness and positive valence. Moving sadness captured an intense experience that involved feelings of sadness and being moved. Nervous sadness was associated with felt anxiety, perceived scariness and negative valence. These interpretations were supported by indirect measures of felt emotion. Experiences of Moving sadness were strongly associated with high trait empathy and emotional contagion, but not with other previously suggested traits such as absorption or nostalgia-proneness. Relaxing sadness and Nervous sadness were not significantly predicted by any of the individual difference variables. The findings are interpreted within a theoretical framework of embodied emotions.