Author name cluster

Kun Bai

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

7 papers

2 author rows

IJCAI Conference 2022 Conference Paper

Contrastive Multi-view Hyperbolic Hierarchical Clustering

Fangfei Lin
Bing Bai
Kun Bai
Yazhou Ren
Peng Zhao
Zenglin Xu

Hierarchical clustering recursively partitions data at an increasingly finer granularity. In real-world applications, multi-view data have become increasingly important. This raises a less investigated problem, i. e. , multi-view hierarchical clustering, to better understand the hierarchical structure of multi-view data. To this end, we propose a novel neural network-based model, namely Contrastive Multi-view Hyperbolic Hierarchical Clustering(CMHHC). It consists of three components, i. e. , multi-view alignment learning, aligned feature similarity learning, and continuous hyperbolic hierarchical clustering. First, we align sample-level representations across multiple views in a contrastive way to capture the view-invariance information. Next, we utilize both the manifold and Euclidean similarities to improve the metric property. Then, we embed the representations into a hyperbolic space and optimize the hyperbolic embeddings via a continuous relaxation of hierarchical clustering loss. Finally, a binary clustering tree is decoded from optimized hyperbolic embeddings. Experimental results on five real-world datasets demonstrate the effectiveness of the proposed method and its components.

PDF Details DOI

AAAI Conference 2022 Conference Paper

Uncertainty-Aware Learning against Label Noise on Imbalanced Datasets

Yingsong Huang
Bing Bai
Shengwei Zhao
Kun Bai
Fei Wang

Learning against label noise is a vital topic to guarantee a reliable performance for deep neural networks. Recent research usually refers to dynamic noise modeling with model output probabilities and loss values, and then separates clean and noisy samples. These methods have gained notable success. However, unlike cherry-picked data, existing approaches often cannot perform well when facing imbalanced datasets, a common scenario in the real world. We thoroughly investigate this phenomenon and point out two major issues that hinder the performance, i. e. , inter-class loss distribution discrepancy and misleading predictions due to uncertainty. The first issue is that existing methods often perform class-agnostic noise modeling. However, loss distributions show a significant discrepancy among classes under class imbalance, and class-agnostic noise modeling can easily get confused with noisy samples and samples in minority classes. The second issue refers to that models may output misleading predictions due to epistemic uncertainty and aleatoric uncertainty, thus existing methods that rely solely on the output probabilities may fail to distinguish confident samples. Inspired by our observations, we propose an Uncertainty-aware Label Correction framework (ULC) to handle label noise on imbalanced datasets. First, we perform epistemic uncertainty-aware classspecific noise modeling to identify trustworthy clean samples and refine/discard highly confident true/corrupted labels. Then, we introduce aleatoric uncertainty in the subsequent learning process to prevent noise accumulation in the label noise modeling process. We conduct experiments on several synthetic and real-world datasets. The results demonstrate the effectiveness of the proposed method, especially on imbalanced datasets.

PDF Details

AAAI Conference 2019 Conference Paper

Compressing Recurrent Neural Networks with Tensor Ring for Action Recognition

Yu Pan
Jing Xu
Maolin Wang
Jinmian Ye
Fei Wang
Kun Bai
Zenglin Xu

Recurrent Neural Networks (RNNs) and their variants, such as Long-Short Term Memory (LSTM) networks, and Gated Recurrent Unit (GRU) networks, have achieved promising performance in sequential data modeling. The hidden layers in RNNs can be regarded as the memory units, which are helpful in storing information in sequential contexts. However, when dealing with high dimensional input data, such as video and text, the input-to-hidden linear transformation in RNNs brings high memory usage and huge computational cost. This makes the training of RNNs very difficult. To address this challenge, we propose a novel compact LSTM model, named as TR-LSTM, by utilizing the low-rank tensor ring decomposition (TRD) to reformulate the input-to-hidden transformation. Compared with other tensor decomposition methods, TR-LSTM is more stable. In addition, TR-LSTM can complete an end-to-end training and also provide a fundamental building block for RNNs in handling large input data. Experiments on real-world action recognition datasets have demonstrated the promising performance of the proposed TR-LSTM compared with the tensor-train LSTM and other state-of-the-art competitors.

PDF Details

IJCAI Conference 2018 Conference Paper

Structured Inference for Recurrent Hidden Semi-markov Model

Hao Liu
Lirong He
Haoli Bai
Bo Dai
Kun Bai
Zenglin Xu

Segmentation and labeling for high dimensional time series is an important yet challenging task in a number of applications, such as behavior understanding and medical diagnosis. Recent advances to model the nonlinear dynamics in such time series data, has suggested to involve recurrent neural networks into Hidden Markov Models. However, this involvement has caused the inference procedure much more complicated, often leading to intractable inference, especially for the discrete variables of segmentation and labeling. To achieve both flexibility and tractability in modeling nonlinear dynamics of discrete variables, we present a structured and stochastic sequential neural network (SSNN), which composes with a generative network and an inference network. In detail, the generative network aims to not only capture the long-term dependencies but also model the uncertainty of the segmentation labels via semi-Markov models. More importantly, for efficient and accurate inference, the proposed bi-directional inference network reparameterizes the categorical segmentation with the Gumbel-Softmax approximation and resorts to the Stochastic Gradient Variational Bayes. We evaluate the proposed model in a number of tasks, including speech modeling, automatic segmentation and labeling in behavior understanding, and sequential multi-objects recognition. Experimental results have demonstrated that our proposed model can achieve significant improvement over the state-of-the-art methods.

PDF Details

ICRA Conference 2011 Conference Paper

Direct field-feedback control for multi-DOF spherical actuators

Kun Bai
Kok-Meng Lee
Shaohui Foong

This paper presents an alternative control strategy for permanent magnet (PM) based spherical actuators capable of multi-DOF precision manipulation. Unlike existing control methods which rely on separate sensing systems, this direct approach utilizes magnetic field measurements for feedback and eliminates the complicated multi-DOF orientation detection in closed-loop control which may cause time-delay and affect system sampling rate. By capitalizing on the rotor magnetic field implicit dependence on orientation, the control law derivation and torque coefficient estimation can be obtained simultaneously and directly from field measurements without explicit determination of the rotor orientation, thereby improving computational efficiency and eliminating error accumulation. The control method is simulated in 2-DOF motion with a CAD model of a spherical actuator.

Details

ICRA Conference 2010 Conference Paper

Magnetic field-based sensing method for spherical joint

Shaohui Foong
Kok-Meng Lee
Kun Bai

This paper presents a sensing method that harnesses the capacity of modern sensors to measure vector fields. This approach directly maps distributed independent field measurements to the instantaneous orientation of a spherical joint embedded with low-cost permanent magnets. Unlike existing methods which require a priori and precise field models, this direct method engages an artificial neural network to associate a collection of measurements to joint orientation. The operation of both bipolar and unipolar single and multi-axis sensors were considered and evaluated experimentally.

Details

ICRA Conference 2009 Conference Paper

Magnetic dipoles for electromagnetic multi-DOF actuator design

Kok-Meng Lee
Jungyoul Lim
Kun Bai

This paper presents a new method for solving the magnetic forces/torques of a multi-DOF spherical actuator that has more controlling inputs than its mechanical DOF. Unlike methods that based on the Lorentz force equation or the Maxwell stress tensor, which require computing the volume or surface integrals to derive the forces, the dipole force method presented here offers the magnetic force solution in closed form. We validate the dipole force model against published experimental data, and demonstrate its application in solving the inverse torque model of a multi-DOF spherical motor, which computes the required set of maximum current inputs for a given design specifications.

Details