AutoDO: Robust AutoAugment for Biased Data with Label Noise via Scalable Probabilistic Implicit Differentiation - Denis Gudovskiy, Luca Rigazio, Shun Ishizaka, Kazuki Kozuka, Sotaro Tsukizawa - CVPR - March 10, 2021. While recent papers evolved in the direction of decreasing policy search complexity, we show that those methods are not robust when applied to biased and noisy data. To overcome these limitations, we reformulate AutoAugment as a generalized automated dataset optimization (AutoDO) task that minimizes the distribution shift between test data and distorted train dataset.
You can also find my articles on my Google Scholar profile.
Deep Active Learning for Biased Datasets via Fisher Kernel Self-Supervision - Denis Gudovskiy, Alec Hodgkinson, Takuya Yamaguchi, Sotaro Tsukizawa - CVPR - March 01, 2020. The main motivation of this paper is to consider a realistic setting for pool-based semi-supervised AL, where the unlabeled collection of train data is biased. We theoretically derive an optimal acquisition function for AL in this setting. It can be formulated as distribution shift minimization between unlabeled train data and weakly-labeled validation dataset.
Smart Home Appliances: Chat with Your Fridge - Denis Gudovskiy, Gyuri Han, Takuya Yamaguchi, Sotaro Tsukizawa - NeurIPS Demo - December 19, 2019. In this paper, we apply state-of-the-art visual reasoning model and demonstrate that it is feasible to ask a smart fridge about its contents and various properties of the food with close-to-natural conversation experience.
Explanation-Based Attention for Semi-Supervised Deep Active Learning - Denis Gudovskiy, Alec Hodgkinson, Takuya Yamaguchi, Sotaro Tsukizawa - LLD @ ICLR - March 20, 2019. We introduce an attention mechanism to improve feature extraction for deep active learning (AL) in the semi-supervised setting.
Explain to Fix: A Framework to Interpret and Correct DNN Object Detector Predictions - Denis Gudovskiy, Alec Hodgkinson, Takuya Yamaguchi, Yasunori Ishii, Sotaro Tsukizawa - SysML @ NeurIPS - November 19, 2018. In this paper, we propose a practical approach to interpret decisions made by a DNN object detector that has fidelity comparable to state-of-the-art methods and sufficient computational efficiency to process large datasets.
DNN Feature Map Compression using Learned Representation over GF(2) - Denis Gudovskiy, Alec Hodgkinson, Luca Rigazio - CEFRL @ ECCV - August 15, 2018. In this paper, we introduce a method to compress intermediate feature maps of deep neural networks (DNNs) to decrease memory storage and bandwidth requirements during inference.
ShiftCNN: Generalized Low-Precision Architecture for Inference of Convolutional Neural Networks - D. Gudovskiy, L. Rigazio - arXiv - June 07, 2017. In this paper we introduce ShiftCNN, a generalized low-precision architecture for inference of multiplierless convolutional neural networks (CNNs).