PTQ-SO: A Scale Optimization-based Approach for Post-training Quantization of Edge Computing.
Kangkang LiuNingjiang ChenPublished in: CSCWD (2024)
Keyphrases
- multiple scales
- optimization algorithm
- stochastic gradient descent
- efficient computation
- training phase
- optimization process
- quantization error
- constrained optimization
- training process
- global optimization
- optimization method
- training examples
- optimization problems
- weighted graph
- decision trees
- optimization methods
- edge information
- scale invariant
- edge detection
- image features