Skip to main navigation
Skip to search
Skip to main content
Tampere University Research Portal Home
Help & FAQ
English
Suomi
Home
Profiles
Research outputs
Research units
Activities
Datasets
Research Infrastructures
Press/Media
Prizes
Search by expertise, name or affiliation
Model compression methods for convolutional neural networks
Henri Lunnikivi
Research output
:
Book/Report
›
Master's thesis
Overview
Fingerprint
Fingerprint
Dive into the research topics of 'Model compression methods for convolutional neural networks'. Together they form a unique fingerprint.
Sort by
Weight
Alphabetically
Keyphrases
Compression Method
100%
Convolutional Neural Network
100%
Model Compression
100%
Latency
50%
Dropout
50%
Computer Vision
50%
Fully Connected Layer
50%
Convolutional Layer
50%
Feature Map
50%
Pruning
25%
Effective Solutions
25%
Mobile Phone
25%
Embedded Platform
25%
Overfitting
25%
Desktop
25%
Memory Bandwidth
25%
Model Size
25%
Extracting Features
25%
Vehicle Classification
25%
Points Reduction
25%
Deep Learning
25%
Processing Resources
25%
CPU-GPU
25%
Mobile IoT
25%
Convolutional Networks
25%
Embedded Target
25%
Vehicle Classifier
25%
Inference Time
25%
IoT Edge Computing
25%
Sparse Formats
25%
Pruned Models
25%
Memory Cache
25%
Sparse Matrix Format
25%
Computer Science
Model Compression
100%
Convolutional Neural Network
100%
Convolutional Layer
50%
Graphics Processing Unit
50%
Computer Vision
50%
Fully Connected Layer
50%
Feature Map
50%
Effective Solution
25%
Embedded Platform
25%
Memory Bandwidth
25%
Regularization
25%
Convolutional Network
25%
Processing Resource
25%
Component Vector
25%
Cache Memory
25%
Internet-Of-Things
25%
Deep Learning Method
25%