Jean-Michel A. Sarr’s research while affiliated with Cheikh Anta Diop University and other places

Efficient learning, that is, learning with small datasets is difficult for current deep learning models. Invariance has been conjectured to be the key for its generalization potential. One of the most used procedure to learn invariant models is data augmentation (DA). Data augmentation can be performed offline by augmenting the data before any training, or it can be performed online during training. However, applying those technique won’t yield better generalization gains every time. We frame this problem as the stability of generalization gains made by invariance inducing techniques. In this study we introduced a new algorithm to train an approximate invariant priors before posterior training of Bayesian Neural Network (BNN). Furthermore, we compared the generalization stability of our invariance inducing algorithm with online DA and offline DA on MNIST and Fashion MNIST with three perturbation processes: rotation, noise, and rotation+noise. Results showed that learning approximate invariant priors requires less exposure to the perturbation process, but it leads BNN to more stable generalization gains during posterior training. Finally, we also show that invariance inducing techniques enhance uncertainty in Bayesian Neural Networks.

Quantitative and qualitative analysis of acoustic backscattered signals from the seabed bottom to the sea surface is used worldwide for fish stocks assessment and marine ecosystem monitoring. Huge amounts of raw data are collected yet require tedious expert labeling. This paper focuses on a case study where the ground truth labels are non-obvious: echograms labeling, which is time-consuming and critical for the quality of fisheries and ecological analysis. We investigate how these tasks can benefit from supervised learning algorithms and demonstrate that convolutional neural networks trained with non-stationary datasets can be used to stress parts of a new dataset needing human expert correction. Further development of this approach paves the way toward a standardization of the labeling process in fisheries acoustics and is a good case study for non-obvious data labeling processes.

Quantitative and qualitative analysis of acoustic backscattered signals from the seabed bottom to the sea surface is used worldwide for fish stocks assessment and marine ecosystem monitoring. Huge amounts of raw data are collected yet require tedious expert labeling. This paper focuses on a case study where the ground truth labels are non-obvious: echograms labeling, which is time-consuming and critical for the quality of fisheries and ecological analysis. We investigate how these tasks can benefit from supervised learning algorithms and demonstrate that convolutional neural networks trained with non-stationary datasets can be used to stress parts of a new dataset needing human expert correction. Further development of this approach paves the way toward a standardization of the labeling process in fisheries acoustics and is a good case study for non-obvious data labeling processes.