An Accuracy-Driven Compression Methodology to Derive Efficient Codebook-Based CNNs.
Flavio PonzinaMiguel Peón QuirósGiovanni AnsaloniDavid AtienzaPublished in: COINS (2022)
Keyphrases
- classification accuracy
- high accuracy
- error rate
- computationally expensive
- vector quantization
- computational complexity
- compression algorithm
- high precision
- neural network
- prediction accuracy
- image representation
- compressed data
- data compression
- cost effective
- image compression
- image classification
- training set
- object recognition
- case study