Since the beginning of 2021, advances in AI research have been revolutionised with the birth of a plethora of deep learning-backed text-to-image models like DALL-E-2, Stable Diffusion, and Midjourney, to name a few. Adding to the list is Google’s Muse, a text-to-image Transformer model that claims to achieve state-of-the-art image generation performance. 
Given the text embedding obtained from a large language model (LLM) that has already been trained, Muse is trained on a masked modelling task in discrete token space. Muse has been trained to predict randomly masked image tokens. Muse asserts to be more effective than pixel-space diffusion models like Imagen and DALL-E 2 since it uses discrete tokens and requires fewer sample iterations. The model generates a zero-shot, mask-free editing for free by iteratively resampling image tokens conditioned on a text prompt.
Check out more about Muse here
Unlike Parti and other autoregressive models, Muse uses parallel decoding. A pre-trained LLM enables fine-grained language understanding, translating to high-fidelity image generation and comprehending visual concepts such as objects, their spatial relationships, pose, cardinality, etc. Additionally, Muse supports inpainting, outpainting, and mask-free editing without the need to modify or invert the model.
With an FID score of 6.06, the 900M parameter model achieves a new SOTA on CC3M. On zero-shot COCO evaluation, the Muse 3B parameter model obtains an FID of 7.88 and a CLIP score of 0.32. 
Model Architecture:

For both the base and super-res Transformer layers, the text encoder creates a text embedding that is used for cross-attention with image tokens. The base model then uses a VQ Tokenizer that generates a 16*16 latent space of tokens after being pre-trained on lower resolution (256*256) images. The cross-entropy loss then learns to predict the masked tokens that have been masked at a variable rate for each sample. After training the base model, the reconstructed lower-resolution tokens and text tokens are then fed into the super-res model. Now the model can predict masked tokens at a higher resolution.
Conference, in-person (Bangalore)
Machine Learning Developers Summit (MLDS) 2023
19-20th Jan, 2023
Early Bird Passes expire on 3rd Feb
Conference, in-person (Bangalore)
Rising 2023 | Women in Tech Conference
16-17th Mar, 2023
Conference, in-person (Bangalore)
Data Engineering Summit (DES) 2023
27-28th Apr, 2023
Conference, in-person (Bangalore)
MachineCon 2023
23rd Jun, 2023
Conference, in-person (Bangalore)
Cypher 2023
20-22nd Sep, 2023
Discover special offers, top stories, upcoming events, and more.
Stay Connected with a larger ecosystem of data science and ML Professionals
For now, the Metaverse’s offerings are as good as a mere rebranding of popular platforms.
AI is playing a vital role in exposing the soft white underbelly of COVID-19. This article looks into the COVID-based AI initiatives of 2021.
The platform is based on a combination of monoliths and microservices, connected using an event-driven model.
The Second-generation AI systems focus on progress where rather than analysing data for diagnosis assistance, prediction, or tailoring therapy, the second-generation platforms help in improving the biological processes.
Intel’s chips in production measure 10 nm while TSMC is producing 5nm already.
AI system developed using computer vision can help surgeons identify the safe zones for dissecting an organ.
Tracking someone after tagging them with Apple’s AirTag is possible given the device’s use of sensors and wireless signals that round up to Apple’s ‘Find My Network’.
As “A Beautiful Mind” completes twenty years, lets take a look at the life and contributions of the legendary mathematician.
Artificial intelligence (AI) has moved from being a niche topic to a dominant technology.
The Rising event aims to bring together some of India’s brightest women data scientists and will be held in person on April 8 in Bengaluru, Karnataka.
© Analytics India Magazine Pvt Ltd & AIM Media House LLC 2023