GAN Cocktail: Mixing GANs Without Dataset Access

Keyphrases

• Generative Models 66%
• Knowledge Model 33%
• Model Merging 100%
• Network Size 33%
• Parameter Space 33%
• Specific Domain 33%
• Target Domain 33%
• Term Model 33%
• Training Data 33%
• Training Model 33%
• Transfer Learning Model 33%
• Two-stage Solution 33%

Computer Science

• Baseline Method 50%
• Generative Model 100%
• Learning Technique 50%
• Parameter Space 50%
• Solution Stage 50%
• Training Data 50%
• Transfer Learning 50%