SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2309.14494 · NeurIPS · 2023

Free-Bloom: Zero-Shot Text-to-Video Generator with LLM Director and LDM Animator

Jingyi Yu, Sibei Yang, Cheng Shi, Hanzhuo Huang, Lan Xu, Yufan Feng

arXiv · PDF · Open in the Atlas

Code that ran

We lifted 11 functions out of this paper's own repositories and ran 5 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.

RepositoryRoleRan
soolab/free-bloom canonical 5 of 11
FunctionStatusWhere it lives
diffusion_step Ran soolab/free-bloom/freebloom/prompt_attention/ptp_utils.py
code served (permissive licence) · get_code("a042902caede448b")
get_noise_pred_single Ran soolab/free-bloom/freebloom/util.py
code served (permissive licence) · get_code("0b088617d5525751")
latent2image Ran soolab/free-bloom/freebloom/prompt_attention/ptp_utils.py
code served (permissive licence) · get_code("e3c3bb87bbe9b4a7")
next_step Ran soolab/free-bloom/freebloom/util.py
code served (permissive licence) · get_code("fb032d6fa7e6196d")
text_under_image Ran soolab/free-bloom/freebloom/prompt_attention/ptp_utils.py
code served (permissive licence) · get_code("45af788f85d9d0c3")
get_down_block Not yet run soolab/free-bloom/freebloom/models/unet_blocks.py
code served (permissive licence) · get_code("a362e5b7a1be4609")
get_matrix Not yet run soolab/free-bloom/freebloom/prompt_attention/seq_aligner.py
code served (permissive licence) · get_code("d167441d997c52ce")
get_matrix Not yet run soolab/free-bloom/freebloom/prompt_attention/seq_aligner.py
code served (permissive licence) · get_code("81db21bf982ca0fe")
get_traceback_matrix Not yet run soolab/free-bloom/freebloom/prompt_attention/seq_aligner.py
code served (permissive licence) · get_code("beb8f6a30678a4d9")
get_up_block Not yet run soolab/free-bloom/freebloom/models/unet_blocks.py
code served (permissive licence) · get_code("01feceecb56a7ffb")
init_prompt Not yet run soolab/free-bloom/freebloom/util.py
code served (permissive licence) · get_code("a0a4d94af120c56f")

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Text-to-video is a rapidly growing research area that aims to generate a semantic, identical, and temporal coherence sequence of frames that accurately align with the input text prompt. This study focuses on zero-shot text-to-video generation considering the data- and cost-efficient. To generate a semantic-coherent video, exhibiting a rich portrayal of temporal semantics such as the whole process of flower blooming rather than a set of "moving images", we propose a novel Free-Bloom pipeline that harnesses large language models (LLMs) as the director to generate a semantic-coherence prompt sequence, while pre-trained latent diffusion models (LDMs) as the animator to generate the high fidelity frames. Furthermore, to ensure temporal and identical coherence while maintaining semantic coherence, we propose a series of annotative modifications to adapting LDMs in the reverse process, including joint noise sampling, step-aware attention shift, and dual-path interpolation. Without any video data and training requirements, Free-Bloom generates vivid and high-quality videos, awe-inspiring in generating complex scenes with semantic meaningful frame sequences. In addition, Free-Bloom is naturally compatible with LDMs-based extensions.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2309.14494")
get_code_for_paper("2309.14494")
have("2309.14494")

Connect an agent — have() is free.