2021/03/01 by A. Asperti, Andrea Asperti, D. Evangelista +6 · 5 citations
Computer Science · Mathematics · #Architecture #Artificial intelligence #Computer science #Data science #Deep learning #FOS: Computer and information sciences #Flexibility (engineering) #Formalism (music) #Generative Adversarial Networks and Image Synthesis #Generative grammar #Generative model #Heuristics #Information retrieval #Machine Learning (cs.LG) #Machine learning #Mathematics #Merge (version control) #Music and Audio Processing #Neural and Evolutionary Computing (cs.NE) #Perspective (graphical) #Representation (politics) #Time Series Analysis and Forecasting #cs.LG #cs.NE
paper · pdf · doi:10.48550/arxiv.2103.01071
published in arXiv (Cornell University) (Cornell University)
arxiv created 2021/03/01 · openalex publication_date 2021/03/01 · arxiv updated 2021/03/02 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Variational AutoEncoders (VAEs) are powerful generative models that merge elements from statistics and information theory with the flexibility offered by deep neural networks to efficiently solve the generation problem for high dimensional data. The key insight of VAEs is to learn the latent distribution of data in such a way that new meaningful samples can be generated from it. This approach led to tremendous research and variations in the architectural design of VAEs, nourishing the recent field of research known as unsupervised representation learning. In this article, we provide a comparative evaluation of some of the most successful, recent variations of VAEs. We particularly focus the analysis on the energetic efficiency of the different models, in the spirit of the so called Green AI, aiming both to reduce the carbon footprint and the financial cost of generative techniques. For each architecture we provide its mathematical formulation, the ideas underlying its design, a detailed model description, a running implementation and quantitative results.