Search Results for "xtream-codes"

We introduce a language modeling approach for text to speech synthesis (TTS). Specifically, we train a neural codec language model (called VALL-E) using discrete codes derived from an off-the-shelf neural audio codec model, and regard TTS as a conditional language modeling task rather than continuous signal regression as in previous work. During the pre-training stage, we scale up the TTS training data to 60K hours of English speech which is hundreds of times larger than existing systems. VALL-E emerges in-context learning capabilities and can be used to synthesize high-quality personalized speech with only a 3-second enrolled recording of an unseen speaker as an acoustic prompt. ...

...It also contains an extension into video and audio generation, using a dual decoder approach. It seems as though a diffusion-based method has taken the new throne for SOTA. However, I will continue on with NUWA, extending it to use multi-headed codes + hierarchical causal transformer. I think that direction is untapped for improving on this line of work. In the paper, they also present a way to condition the video generation based on segmentation mask(s). You can easily do this as well, given you train a VQGanVAE on the sketches beforehand. Then, you will use NUWASketch instead of NUWA, which can accept the sketch VAE as a reference. ...

A simple PyTorch Implementation of Generative Adversarial Networks, focusing on anime face drawing. The images are generated from a DCGAN model trained on 143,000 anime character faces for 100 epochs. Manipulating latent codes enables the transition from images in the first row to the last row. The images are not clean, some outliers can be observed, which degrades the quality of the generated images. Anime-style images of 126 tags are collected from danbooru.donmai.us using the crawler tool gallery-dl. The images are then processed by an anime face detector python-anime face. ...

...You can use it by importing the HierarchicalCausalTransformer. For autoregressive (AR) modeling of high-resolution images, vector quantization (VQ) represents an image as a sequence of discrete codes. A short sequence length is important for an AR model to reduce its computational costs to consider long-range interactions of codes. However, we postulate that previous VQ cannot shorten the code sequence and generate high-fidelity images together in terms of the rate-distortion trade-off.

...The paper proposes a new discrete latent representation named Memcodes, which can be succinctly described as a type of multi-head hard-attention to learned memory (codebook) key/values. They claim the need for less codes and smaller codebook dimensions in order to achieve better reconstructions.