Heterogeneous Multi-Task Learning With Expert Diversity.

Predicting multiple heterogeneous biological and medical targets is a challenge for traditional deep learning models. In contrast to single-task learning, in which a separate model is trained for each target, multi-task learning (MTL) optimizes a single model to predict multiple related targets simultaneously. To address this challenge, we propose the Multi-gate Mixture-of-Experts with Exclusivity (MMoEEx). Our work aims to tackle the heterogeneous MTL setting, in which the same model optimizes multiple tasks with different characteristics. Such a scenario can overwhelm current MTL approaches due to the challenges in balancing shared and task-specific representations and the need to optimize tasks with competing optimization paths. Our method makes two key contributions: first, we introduce an approach to induce more diversity among experts, thus creating representations more suitable for highly imbalanced and heterogenous MTL learning; second, we adopt a two-step optimization (Finn et al., 2017 and Lee et al., 2020) approach to balancing the tasks at the gradient level. We validate our method on three MTL benchmark datasets, including UCI-Census-income dataset, Medical Information Mart for Intensive Care (MIMIC-III) and PubChem BioAssay (PCBA).

Deep Neural Network Compression by In-Parallel Pruning-Quantization.

Deep neural networks enable state-of-the-art accuracy on visual recognition tasks such as image classification and object detection. However, modern networks contain millions of learned connections, and the current trend is towards deeper and more densely connected architectures. This poses a challenge to the deployment of state-of-the-art networks on resource-constrained systems, such as smartphones or mobile robots. In general, a more efficient utilization of computation resources would assist in deployment scenarios from embedded platforms to computing clusters running ensembles of networks. In this paper, we propose a deep network compression algorithm that performs weight pruning and quantization jointly, and in parallel with fine-tuning. Our approach takes advantage of the complementary nature of pruning and quantization and recovers from premature pruning errors, which is not possible with two-stage approaches. In experiments on ImageNet, CLIP-Q (Compression Learning by In-Parallel Pruning-Quantization) improves the state-of-the-art in network compression on AlexNet, VGGNet, GoogLeNet, and ResNet. We additionally demonstrate that CLIP-Q is complementary to efficient network architecture design by compressing MobileNet and ShuffleNet, and that CLIP-Q generalizes beyond convolutional networks by compressing a memory network for visual question answering.

A decoupled approach to illumination-robust optical flow estimation.

Despite continuous improvements in optical flow in the last three decades, the ability for optical flow algorithms to handle illumination variation is still an unsolved challenge. To improve the ability to interpret apparent object motion in video containing illumination variation, an illumination-robust optical flow method is designed. This method decouples brightness into reflectance and illumination components using a stochastic technique; reflectance is given higher weight to ensure robustness against illumination, which is suppressed. Illumination experiments using the Middlebury and University of Oulu databases demonstrate the decoupled method's improvement when compared with state-of-the-art. In addition, a novel technique is implemented to visualize optical flow output, which is especially useful to compare different optical flow methods in the absence of the ground truth.