S.-H. Gary Chan
YOU?
Author Swipe
View article: MTM: A Multi-Scale Token Mixing Transformer for Irregular Multivariate Time Series Classification
MTM: A Multi-Scale Token Mixing Transformer for Irregular Multivariate Time Series Classification Open
Irregular multivariate time series (IMTS) is characterized by the lack of synchronized observations across its different channels. In this paper, we point out that this channel-wise asynchrony can lead to poor channel-wise modeling of exis…
View article: RateCount: Learning-Free Device Counting by Wi-Fi Probe Listening
RateCount: Learning-Free Device Counting by Wi-Fi Probe Listening Open
A Wi-Fi-enabled device, or simply Wi-Fi device, sporadically broadcasts probe request frames (PRFs) to discover nearby access points (APs), whether connected to an AP or not. To protect user privacy, unconnected devices often randomize the…
View article: Advancements in PET bottle plastic slitting devices for raw material 3D printing filament production
Advancements in PET bottle plastic slitting devices for raw material 3D printing filament production Open
The world is facing a significant environmental challenge due to the accumulation of plastic waste, especially PET (Polyethylene Terephthalate) bottles. Indonesia is no exception to this problem. Every year, millions of tons of PET bottles…
View article: Graph-based Fingerprint Update Using Unlabelled WiFi Signals
Graph-based Fingerprint Update Using Unlabelled WiFi Signals Open
WiFi received signal strength (RSS) environment evolves over time due to the movement of access points (APs), AP power adjustment, installation and removal of APs, etc. We study how to effectively update an existing database of fingerprint…
View article: DLiGRU-X: Efficient X-Vector-Based Embeddings for Small-Footprint Keyword Spotting System
DLiGRU-X: Efficient X-Vector-Based Embeddings for Small-Footprint Keyword Spotting System Open
Deployment of deep learning-based speech processing models for real-world applications on devices with limited processing capacity and memory constraints poses significant challenges. This paper introduces an enhanced deep learning model b…
View article: SnapGen: Taming High-Resolution Text-to-Image Models for Mobile Devices with Efficient Architectures and Training
SnapGen: Taming High-Resolution Text-to-Image Models for Mobile Devices with Efficient Architectures and Training Open
Existing text-to-image (T2I) diffusion models face several limitations, including large model sizes, slow runtime, and low-quality generation on mobile devices. This paper aims to address all of these challenges by developing an extremely …
View article: M$^3$-Impute: Mask-guided Representation Learning for Missing Value Imputation
M$^3$-Impute: Mask-guided Representation Learning for Missing Value Imputation Open
Missing values are a common problem that poses significant challenges to data analysis and machine learning. This problem necessitates the development of an effective imputation method to fill in the missing values accurately, thereby enha…
View article: Revisiting Referring Expression Comprehension Evaluation in the Era of Large Multimodal Models
Revisiting Referring Expression Comprehension Evaluation in the Era of Large Multimodal Models Open
Referring expression comprehension (REC) involves localizing a target instance based on a textual description. Recent advancements in REC have been driven by large multimodal models (LMMs) like CogVLM, which achieved 92.44% accuracy on Ref…
View article: Experiences of Deploying a Citywide Crowdsourcing Platform to Search for Missing People with Dementia
Experiences of Deploying a Citywide Crowdsourcing Platform to Search for Missing People with Dementia Open
People with Dementia (PwD) suffer from a high risk of getting lost due to their cognitive deterioration, leading to potential safety hazards and significant search efforts. In this paper, we propose DEmentia Caring System (DECS), an effect…
View article: Elevator, Escalator, or Neither? Classifying Conveyor State Using Smartphone under Arbitrary Pedestrian Behavior
Elevator, Escalator, or Neither? Classifying Conveyor State Using Smartphone under Arbitrary Pedestrian Behavior Open
Knowing a pedestrian's conveyor state of ''elevator,'' ''escalator,'' or ''neither'' is fundamental to many applications such as indoor navigation and people flow management. Previous studies on classifying the conveyor state often rely on…
View article: Single Domain Generalization for Crowd Counting
Single Domain Generalization for Crowd Counting Open
Due to its promising results, density map regression has been widely employed for image-based crowd counting. The approach, however, often suffers from severe performance degradation when tested on data from unseen scenarios, the so-called…
View article: StableKD: Breaking Inter-block Optimization Entanglement for Stable Knowledge Distillation
StableKD: Breaking Inter-block Optimization Entanglement for Stable Knowledge Distillation Open
Knowledge distillation (KD) has been recognized as an effective tool to compress and accelerate models. However, current KD approaches generally suffer from an accuracy drop and/or an excruciatingly long distillation process. In this paper…
View article: Target-agnostic Source-free Domain Adaptation for Regression Tasks
Target-agnostic Source-free Domain Adaptation for Regression Tasks Open
Unsupervised domain adaptation (UDA) seeks to bridge the domain gap between the target and source using unlabeled target data. Source-free UDA removes the requirement for labeled source data at the target to preserve data privacy and stora…
View article: A Multi-Scale Decomposition MLP-Mixer for Time Series Analysis
A Multi-Scale Decomposition MLP-Mixer for Time Series Analysis Open
Time series data, including univariate and multivariate ones, are characterized by unique composition and complex multi-scale temporal variations. They often require special consideration of decomposition and multi-scale modeling to analyz…
View article: Improving Out-of-Distribution Robustness of Classifiers via Generative Interpolation
Improving Out-of-Distribution Robustness of Classifiers via Generative Interpolation Open
Deep neural networks achieve superior performance for learning from independent and identically distributed (i.i.d.) data. However, their performance deteriorates significantly when handling out-of-distribution (OoD) data, where the traini…
View article: FIS-ONE: Floor Identification System with One Label for Crowdsourced RF Signals
FIS-ONE: Floor Identification System with One Label for Crowdsourced RF Signals Open
Floor labels of crowdsourced RF signals are crucial for many smart-city applications, such as multi-floor indoor localization, geofencing, and robot surveillance. To build a prediction model to identify the floor number of a new RF signal …
View article: Agent-Based Modelling for Real-World Stock Markets under Behavioral Economic Principles
Agent-Based Modelling for Real-World Stock Markets under Behavioral Economic Principles Open
The reproduction of realistic dynamics in financial markets is of great significance, as it enhances our understanding of market evolution beyond other physical processes, and facilitates the development and backtesting of investment strat…
View article: Leto: Crowdsourced Radio Map Construction With Learned Topology and a Few Landmarks
Leto: Crowdsourced Radio Map Construction With Learned Topology and a Few Landmarks Open
Existing crowdsourced indoor positioning systems (CIPSs) usually require prior knowledge about the site and a tedious calibration process. Moreover, they may require a large number of landmarks while ignoring the topology information that …
View article: Run, Don't Walk: Chasing Higher FLOPS for Faster Neural Networks
Run, Don't Walk: Chasing Higher FLOPS for Faster Neural Networks Open
To design fast neural networks, many works have been focusing on reducing the number of floating-point operations (FLOPs). We observe that such reduction in FLOPs, however, does not necessarily lead to a similar level of reduction in laten…
View article: Semi-supervised Learning with Network Embedding on Ambient RF Signals for Geofencing Services
Semi-supervised Learning with Network Embedding on Ambient RF Signals for Geofencing Services Open
In applications such as elderly care, dementia anti-wandering and pandemic control, it is important to ensure that people are within a predefined area for their safety and well-being. We propose GEM, a practical, semi-supervised Geofencing…
View article: GRAFICS: Graph Embedding-based Floor Identification Using Crowdsourced RF Signals
GRAFICS: Graph Embedding-based Floor Identification Using Crowdsourced RF Signals Open
We study the problem of floor identification for radiofrequency (RF) signal samples obtained in a crowdsourced manner, where the signal samples are highly heterogeneous and most samples lack their floor labels. We propose GRAFICS, a graph …
View article: GRAFICS: Graph Embedding-based Floor Identification Using Crowdsourced RF Signals
GRAFICS: Graph Embedding-based Floor Identification Using Crowdsourced RF Signals Open
We study the problem of floor identification for radiofrequency (RF) signal samples obtained in a crowdsourced manner, where the signal samples are highly heterogeneous and most samples lack their floor labels. We propose GRAFICS, a graph …
View article: TVConv: Efficient Translation Variant Convolution for Layout-aware Visual Processing
TVConv: Efficient Translation Variant Convolution for Layout-aware Visual Processing Open
As convolution has empowered many smart applications, dynamic convolution further equips it with the ability to adapt to diverse inputs. However, the static and dynamic convolutions are either layout-agnostic or computation-heavy, making i…
View article: Tackling Multipath and Biased Training Data for IMU-Assisted BLE Proximity Detection
Tackling Multipath and Biased Training Data for IMU-Assisted BLE Proximity Detection Open
Proximity detection is to determine whether an IoT receiver is within a certain distance from a signal transmitter. Due to its low cost and high popularity, Bluetooth low energy (BLE) has been used to detect proximity based on the received…
View article: A Data-Driven Spatial-Temporal Graph Neural Network for Docked Bike Prediction
A Data-Driven Spatial-Temporal Graph Neural Network for Docked Bike Prediction Open
Docked bike systems have been widely deployed in many cities around the world. To the service provider, predicting the demand and supply of bikes at any station is crucial to offering the best service quality. The docked bike prediction pr…
View article: TVConv: Efficient Translation Variant Convolution for Layout-aware Visual Processing
TVConv: Efficient Translation Variant Convolution for Layout-aware Visual Processing Open
As convolution has empowered many smart applications, dynamic convolution further equips it with the ability to adapt to diverse inputs. However, the static and dynamic convolutions are either layout-agnostic or computation-heavy, making i…
View article: Tackling Multipath and Biased Training Data for IMU-Assisted BLE Proximity Detection
Tackling Multipath and Biased Training Data for IMU-Assisted BLE Proximity Detection Open
Proximity detection is to determine whether an IoT receiver is within a certain distance from a signal transmitter. Due to its low cost and high popularity, Bluetooth low energy (BLE) has been used to detect proximity based on the received…
View article: A Lightweight and Accurate Spatial-Temporal Transformer for Traffic Forecasting
A Lightweight and Accurate Spatial-Temporal Transformer for Traffic Forecasting Open
We study the forecasting problem for traffic with dynamic, possibly periodical, and joint spatial-temporal dependency between regions. Given the aggregated inflow and outflow traffic of regions in a city from time slots 0 to t-1, we predic…
View article: NAS-OoD: Neural Architecture Search for Out-of-Distribution Generalization
NAS-OoD: Neural Architecture Search for Out-of-Distribution Generalization Open
Recent advances on Out-of-Distribution (OoD) generalization reveal the robustness of deep learning models against distribution shifts. However, existing works focus on OoD algorithms, such as invariant risk minimization, domain generalizat…