Towards a Definition of Disentangled Representations (2018, background): “This is the motivation for the work on disentangled representation learning, which differs from other dimensionality reduction approaches through its explicit aim to learn a representation that axis aligns with the generative factors of the data [10, 67, 25].”
Exponential expressivity in deep neural networks through transient chaos (2016, neutral): “The second piece of intuition, which has captured the imagination of machine learning [7] and neuroscience [8] alike, is that deep neural networks can disentangle highly curved manifolds in input space into flattened manifolds in hidden space.”
Encoding and decoding in fMRI (2010, background): “Finally, the mapping from pictures (the input space) to category labels (the feature space) is highly nonlinear (DiCarlo and Cox, 2007), while the mapping from activity space to feature space is supplied by the linear classifier.”
Image webs: Computing and exploiting connectivity in image collections (2010, background): “As we briefly survey in Section 2, considerable research has already been devoted to exploiting match links to detect images of popular landmarks [25] and to annotate them [11], or for 3-”
Deep Hierarchies in the Primate Visual Cortex: What Can We Learn for Computer Vision? (2013, background): “…Ettinger [42], Geman et al. [59], [60], Mel and Fiser [114], Amit [1], [2], Hawkins [69], Fidler et al. [45], Scalzo and Piater [152], Ullman and Epshtein [191], DiCarlo and Cox [31], Ommer and Buhmann [126], Serre and Poggio [156], Pugeault et al. [207] and Rodríguez-Sánchez and Tsotsos [143].”