• Publicationsclose
• Other research productsclose
• 2019-2023close
• Preprint close
• AE close

We propose a data-driven approach for deep convolutional neural network compression that achieves high accuracy with high throughput and low memory requirements. Current network compression methods either find a low-rank factorization of the features that requires more memory, or select only a subset of features by pruning entire filter channels. We propose the Cascaded Projection (CaP) compression method that projects the output and input filter channels of successive layers to a unified low dimensional space based on a low-rank projection. We optimize the projection to minimize classification loss and the difference between the next layer's features in the compressed and uncompressed networks. To solve this non-convex optimization problem we propose a new optimization method of a proxy matrix using backpropagation and Stochastic Gradient Descent (SGD) with geometric constraints. Our cascaded projection approach leads to improvements in all critical areas of network compression: high accuracy, low memory consumption, low parameter count and high processing speed. The proposed CaP method demonstrates state-of-the-art results compressing VGG16 and ResNet networks with over 4x reduction in the number of computations and excellent performance in top-5 accuracy on the ImageNet dataset before and after fine-tuning.

Prototype-based meta-learning has emerged as a powerful technique for addressing few-shot learning challenges. However, estimating a deterministic prototype using a simple average function from a limited number of examples remains a fragile process. To overcome this limitation, we introduce ProtoDiff, a novel framework that leverages a task-guided diffusion model during the meta-training phase to gradually generate prototypes, thereby providing efficient class representations. Specifically, a set of prototypes is optimized to achieve per-task prototype overfitting, enabling accurately obtaining the overfitted prototypes for individual tasks. Furthermore, we introduce a task-guided diffusion process within the prototype space, enabling the meta-learning of a generative process that transitions from a vanilla prototype to an overfitted prototype. ProtoDiff gradually generates task-specific prototypes from random noise during the meta-test stage, conditioned on the limited samples available for the new task. Furthermore, to expedite training and enhance ProtoDiff's performance, we propose the utilization of residual prototype learning, which leverages the sparsity of the residual prototype. We conduct thorough ablation studies to demonstrate its ability to accurately capture the underlying prototype distribution and enhance generalization. The new state-of-the-art performance on within-domain, cross-domain, and few-task few-shot classification further substantiates the benefit of ProtoDiff. Comment: Under review

Person re-identification has achieved great progress with deep convolutional neural networks. However, most previous methods focus on learning individual appearance feature embedding, and it is hard for the models to handle difficult situations with different illumination, large pose variance and occlusion. In this work, we take a step further and consider employing context information for person search. For a probe-gallery pair, we first propose a contextual instance expansion module, which employs a relative attention module to search and filter useful context information in the scene. We also build a graph learning framework to effectively employ context pairs to update target similarity. These two modules are built on top of a joint detection and instance feature learning framework, which improves the discriminativeness of the learned features. The proposed framework achieves state-of-the-art performance on two widely used person search datasets. Comment: To appear in CVPR 2019

Neural memory enables fast adaptation to new tasks with just a few training samples. Existing memory models store features only from the single last layer, which does not generalize well in presence of a domain shift between training and test distributions. Rather than relying on a flat memory, we propose a hierarchical alternative that stores features at different semantic levels. We introduce a hierarchical prototype model, where each level of the prototype fetches corresponding information from the hierarchical memory. The model is endowed with the ability to flexibly rely on features at different semantic levels if the domain shift circumstances so demand. We meta-learn the model by a newly derived hierarchical variational inference framework, where hierarchical memory and prototypes are jointly optimized. To explore and exploit the importance of different semantic levels, we further propose to learn the weights associated with the prototype at each level in a data-driven way, which enables the model to adaptively choose the most generalizable features. We conduct thorough ablation studies to demonstrate the effectiveness of each component in our model. The new state-of-the-art performance on cross-domain and competitive performance on traditional few-shot classification further substantiates the benefit of hierarchical variational memory. Comment: 17 pages, 5 figures

While there are many studies on pareidolia in healthy individuals and patients with schizophrenia, to our knowledge, there are no prior studies on pareidolia in patients with bipolar disorder. Accordingly, in this study, we, for the first time, measured pareidolia in patients with bipolar disorder (N = 50), and compared that to patients with schizophrenia (N = 50) and healthy controls (N = 50). We have used (a) the scene test, which consists of 10 blurred images of natural scenes that was previously found to produce illusory face responses and (b) the noise test which had 32 black and white images consisting of visual noise and 8 images depicting human faces; participants indicated whether a face was present on these images and to point to the location where they saw the face. Illusory responses were defined as answers when observers falsely identified objects that were not on the images in the scene task (maximum illusory score: 10), and the number of noise images in which they reported the presence of a face (maximum illusory score: 32). Further, we also calculated the total pareidolia score for each task (the sum number of images with illusory responses in the scene and noise tests). The responses were scored by two independent raters with an excellent congruence (kappa > 0.9). Our results show that schizophrenia patients scored higher on pareidolia measures than both healthy controls and patients with bipolar disorder. Our findings are agreement with prior findings on more impaired cognitive processes in schizophrenia than in bipolar patients.

What rhetorics run throughout radical discourse, and why do some gain prominence over others? The scholarship on radicalism largely portrays radical discourse as opposition to powerful ideas and enemies, but radicals often evince great interest in personal and local concerns. To shed light on how radicals use and adopt rhetoric, we analyze an original corpus of more than 23,000 pages produced by Afghan radical groups between 1979 and 2001 using a novel computational abductive approach. We first identify how radicalism not only attacks dominant ideas, actors, and institutions using a rhetoric of subversion, but also how it can use a rhetoric of reversion to urge intimate transformations in morals and behavior. Next, we find evidence that radicals’ networks of support affect the rhetorical mixture they espouse, due to social ties drawing radicals into encounters with backers’ social domains. Our study advances a relational understanding of radical discourse, while also showing how a combination of computational and abductive methods can help theorize and analyze discourses of contention.

A characteristic feature of the 3d plaquette Ising model is its planar subsystem symmetry. The quantum version of this model has been shown to be related via a duality to the X-Cube model, which has been paradigmatic in the new and rapidly developing field of fractons. The relation between the 3d plaquette Ising and the X-Cube model is similar to that between the 2d quantum transverse spin Ising model and the Toric Code. Gauging the global symmetry in the case of the 2d Ising model and considering the gauge invariant sector of the high temperature phase leads to the Toric Code, whereas gauging the subsystem symmetry of the 3d quantum transverse spin plaquette Ising model leads to the X-Cube model. A non-standard dual formulation of the 3d plaquette Ising model which utilises three flavours of spins has recently been discussed in the context of dualising the fracton-free sector of the X-Cube model. In this paper we investigate the classical spin version of this non-standard dual Hamiltonian and discuss its properties in relation to the more familiar Ashkin-Teller-like dual and further related dual formulations involving both link and vertex spins and non-Ising spins. Comment: Reviews results in arXiv:1106.0325 and arXiv:1106.4664 in light of more recent simulations and fracton literature. Published in special issue of Entropy dedicated to the memory of Professor Ian Campbell