- San Francisco Bay Area
-
00:30
(UTC -08:00) - https://halosnoopy.github.io/HuangLin.github.io/
Highlights
- Pro
Pinned Loading
-
A-YOLOv5s-based-Steel-Surface-Defects-Detector
A-YOLOv5s-based-Steel-Surface-Defects-Detector PublicA computer vision project implementing the YOLOv5s model for detecting and classifying steel surface defects using the NEU-DET dataset. Includes dataset preparation, training, evaluation, and GUI s…
Python 1
-
CNN-Based-Indoor-Positioning
CNN-Based-Indoor-Positioning PublicA Machine Learning project that uses convolutional neural networks (CNNs) to predict indoor areas, floors, and buildings based on Wi-Fi RSS fingerprints from the UJIIndoorLoc dataset.
MATLAB 1
-
Group-Matching-Method-for-Indoor-Localization
Group-Matching-Method-for-Indoor-Localization PublicA region-level filtering method for Wi-Fi indoor localization that identifies dominant Access Points to reduce the search space before applying machine learning algorithms for position estimation.
Jupyter Notebook 1
-
LSM-for-ToA-Wireless-Positioning
LSM-for-ToA-Wireless-Positioning PublicA MATLAB project demonstrating the use of the LSM for solving trilateration equations in ToA based wireless positioning. Includes simulation, localization estimation, RMSE analysis, and visualizati…
MATLAB 1
-
LSTM-MLP-Knowledge-Distillation-for-Indoor-Localization
LSTM-MLP-Knowledge-Distillation-for-Indoor-Localization PublicImplements knowledge distillation between an LSTM teacher and an MLP student for Wi-Fi indoor localization. The LSTM captures temporal features from signal sequences, while the MLP learns from the …
Jupyter Notebook 1
-
TSFA-for-Indoor-Localization
TSFA-for-Indoor-Localization PublicA two-stage Wi-Fi fingerprinting method for indoor localization that uses transformed signal patterns and similarity filtering to refine candidate reference points before adaptive position estimation.
Jupyter Notebook 1
If the problem persists, check the GitHub status page or contact support.