Irida Labs (Greece)

The increasing global demand for sustainable seafood calls for innovative solutions that reduce reliance on marine resources. Organic aquaculture, with its focus on environmental sustainability and animal welfare, represents a promising avenue. However, one of the major constraints to its expansion is the limited availability of certified organic feed ingredients. In this study, we evaluated the feasibility of replacing fishmeal (FM) in the diets of meagre (Argyrosomus regius), a promising aquaculture species, by increasing the inclusion of organic pea protein meal (0%, 7%, 18% and 27.5%), while supplementing all pea-protein enriched diets with fixed levels of functional organic ingredients (5% yeast and 2% brown seaweed). Four isoenergetic diets—C (control), PEA 7%, PEA 18% and PEA 27%—were administered to triplicate groups of 40 juvenile meagre over 85 days. We assessed growth performance, feed utilisation, body composition, plasma biochemistry, and gut microbiota composition. All pea protein inclusion levels (7%–27.5%) showed comparable growth to the control diet, with no significant differences in final body weight (PEA 7%: 173 ± 22.5 g; PEA 18%: 189 ± 11.7 g; PEA 27%: 178 ± 23.7 g vs. control: 171 ± 10.0 g) or feed conversion ratio (FCR), demonstrating successful FM replacement up to 27.5%. The PEA 18% group showed superior lipid utilisation with 109% ± 7.21% gross lipid efficiency (GLE; vs. 75.4% ± 10.8% in PEA 7% and 85.3% ± 5.47% in control) and optimal protein metabolism (albumin [ALB]: 0.71 ± 0.07 g/dL vs. control: 0.62 ± 0.05), indicating enhanced nutrient utilisation at intermediate inclusion. High pea inclusion (27.5%) significantly increased beneficial Lactobacillus (8.74% ± 3.30% vs. control: 4.92% ± 3.96%) while maintaining microbial diversity, suggesting pre-biotic effects without dysbiosis. Overall, the results demonstrated that organic pea protein as well as organic seaweed and yeast can be successfully incorporated into meagre diets without compromising growth, feed efficiency, or fish health. These findings highlight the potential of organic pea meal as a sustainable protein alternative to FM in organic aquafeeds for meagre.

Despite the success of hand-crafted features in computer visioning for many years, nowadays, this has been replaced by end-to-end learnable features that are extracted from deep convolutional neural networks (CNNs). Whilst CNNs can learn robust features directly from image pixels, they require large amounts of samples and extreme augmentations. On the contrary, hand-crafted features, like SIFT, exhibit several interesting properties as they can provide local rotation invariance. In this work, a novel scheme combining the strengths of SIFT descriptors with CNNs, namely SIFT-CNN, is presented. Given a single-channel image, one SIFT descriptor is computed for every pixel, and thus, every pixel is represented as an M-dimensional histogram, which ultimately results in an M-channel image. Thus, the SIFT image is generated from the SIFT descriptors for all the pixels in a single-channel image, while at the same time, the original spatial size is preserved. Next, a CNN is trained to utilize these M-channel images as inputs by operating directly on the multiscale SIFT images with the regular convolution processes. Since these images incorporate spatial relations between the histograms of the SIFT descriptors, the CNN is guided to learn features from local gradient information of images that otherwise can be neglected. In this manner, the SIFT-CNN implicitly acquires a local rotation invariance property, which is desired for problems where local areas within the image can be rotated without affecting the overall classification result of the respective image. Some of these problems refer to indirect immunofluorescence (IIF) cell image classification, ground-based all-sky image-cloud classification and human lip-reading classification. The results for the popular datasets related to the three different aforementioned problems indicate that the proposed SIFT-CNN can improve the performance and surpasses the corresponding CNNs trained directly on pixel values in various challenging tasks due to its robustness in local rotations. Our findings highlight the importance of the input image representation in the overall efficiency of a data-driven system.

The use of Deep Learning (DL) algorithms is increasingly evolving in many application domains. Despite the rapid growing of algorithm size and complexity, performing DL inference at the edge is becoming a clear trend to cope with low latency, privacy and bandwidth constraints. Nevertheless, traditional implementation on low-energy computing nodes often requires experience-based manual intervention and trial-and-error iterations to get to a functional and effective solution. This work presents a computer-aided design (CAD) support for effective implementation of DL algorithms on embedded systems, aiming at automating different design steps and reducing cost. The proposed tool flow comprises capabilities to consider architecture-and hardware-related variables at very early stages of the development process, from pre-training hyperparameter optimization and algorithm configuration to deployment, and to adequately address security, power efficiency and adaptivity requirements. This paper also presents some preliminary results obtained by the first implementation of the optimization techniques supported by the tool flow.