Dicetrack: Lightweight Dice Classification on Resource-Constrained Platforms with Optimized Deep Learning Models.
Christophe El ZeinatyGlenn HerrouWassim HamidoucheDaniel MénardPublished in: ICASSP (2024)
Keyphrases
- lightweight
- resource constrained
- learning models
- wireless sensor networks
- classification models
- rfid tags
- machine learning
- machine learning algorithms
- development environments
- learning algorithm
- loss function
- resource constraints
- image classification
- classification accuracy
- decision trees
- text classification
- learning tasks
- feature extraction
- multipath
- sensor networks
- feature space
- semi supervised learning
- support vector
- sensor nodes
- training data
- supervised learning
- feature selection
- data sets