AMD-Hummingbird: Towards an Efficient Text-to-Video Model
- URL: http://arxiv.org/abs/2503.18559v2
- Date: Tue, 25 Mar 2025 02:43:16 GMT
- Title: AMD-Hummingbird: Towards an Efficient Text-to-Video Model
- Authors: Takashi Isobe, He Cui, Dong Zhou, Mengmeng Ge, Dong Li, Emad Barsoum,
- Abstract summary: Text-to-Video (T2V) generation has attracted significant attention for its ability to synthesize realistic videos from textual descriptions.<n>Most prior work prioritizes visual fidelity while overlooking the need for smaller, more efficient models suitable for real-world deployment.<n>We propose a lightweight T2V framework, termed Hummingbird, which prunes existing models and enhances visual quality through visual feedback learning.
- Score: 12.09360569154206
- License: http://creativecommons.org/publicdomain/zero/1.0/
- Abstract: Text-to-Video (T2V) generation has attracted significant attention for its ability to synthesize realistic videos from textual descriptions. However, existing models struggle to balance computational efficiency and high visual quality, particularly on resource-limited devices, e.g.,iGPUs and mobile phones. Most prior work prioritizes visual fidelity while overlooking the need for smaller, more efficient models suitable for real-world deployment. To address this challenge, we propose a lightweight T2V framework, termed Hummingbird, which prunes existing models and enhances visual quality through visual feedback learning. Our approach reduces the size of the U-Net from 1.4 billion to 0.7 billion parameters, significantly improving efficiency while preserving high-quality video generation. Additionally, we introduce a novel data processing pipeline that leverages Large Language Models (LLMs) and Video Quality Assessment (VQA) models to enhance the quality of both text prompts and video data. To support user-driven training and style customization, we publicly release the full training code, including data processing and model training. Extensive experiments show that our method achieves a 31X speedup compared to state-of-the-art models such as VideoCrafter2, while also attaining the highest overall score on VBench. Moreover, our method supports the generation of videos with up to 26 frames, addressing the limitations of existing U-Net-based methods in long video generation. Notably, the entire training process requires only four GPUs, yet delivers performance competitive with existing leading methods. Hummingbird presents a practical and efficient solution for T2V generation, combining high performance, scalability, and flexibility for real-world applications.
Related papers
- ARC-Hunyuan-Video-7B: Structured Video Comprehension of Real-World Shorts [56.75723197779384]
ARC-Hunyuan-Video is a multimodal model that processes visual, audio, and textual signals end-to-end for structured comprehension.<n>Our model is capable of multi-granularity timestamped video captioning and summarization, open-ended video question answering, temporal video grounding, and video reasoning.
arXiv Detail & Related papers (2025-07-28T15:52:36Z) - Encapsulated Composition of Text-to-Image and Text-to-Video Models for High-Quality Video Synthesis [14.980220974022982]
We introduce EVS, a training-free Encapsulated Video Synthesizer that composes T2I and T2V models to enhance both visual fidelity and motion smoothness.<n>Our approach utilizes a well-trained diffusion-based T2I model to refine low-quality video frames.<n>We also employ T2V backbones to ensure consistent motion dynamics.
arXiv Detail & Related papers (2025-07-18T08:59:02Z) - GigaVideo-1: Advancing Video Generation via Automatic Feedback with 4 GPU-Hours Fine-Tuning [38.85475786611648]
GigaVideo-1 is an efficient fine-tuning framework that advances video generation without additional human supervision.<n>We focus on two key aspects of the fine-tuning process: data and optimization.<n>Experiments show that GigaVideo-1 consistently improves performance on almost all the dimensions with an average gain of about 4% using only 4 GPU-hours.
arXiv Detail & Related papers (2025-06-12T12:25:37Z) - Seedance 1.0: Exploring the Boundaries of Video Generation Models [71.26796999246068]
Seedance 1.0 is a high-performance and inference-efficient video foundation generation model.<n>It integrates multi-source curation data augmented with precision and meaningful video captioning.<n>Seedance 1.0 can generate a 5-second video at 1080p resolution only with 41.4 seconds ( NVIDIA-L20)
arXiv Detail & Related papers (2025-06-10T17:56:11Z) - ContentV: Efficient Training of Video Generation Models with Limited Compute [16.722018026516867]
ContentV is a text-to-video model that generates diverse, high-quality videos across multiple resolutions and durations from text prompts.<n>It achieves state-of-the-art performance (85.14 on VBench) after training on 256 x 64GB Neural Processing Units (NPUs) for merely four weeks.
arXiv Detail & Related papers (2025-06-05T17:59:54Z) - Pretrained Image-Text Models are Secretly Video Captioners [38.66202065611397]
We find that an image-based model can be repurposed to outperform several specialised video captioning systems.<n>Our adapted model demonstrates top tier performance on major benchmarks, ranking 2nd on MSRVTT and MSVD, and 3rd on VATEX.<n>From a resource optimization perspective, this video captioning study focuses on three fundamental factors: optimizing model scale, maximizing data efficiency, and incorporating reinforcement learning.
arXiv Detail & Related papers (2025-02-19T01:53:03Z) - Step-Video-T2V Technical Report: The Practice, Challenges, and Future of Video Foundation Model [133.01510927611452]
We present Step-Video-T2V, a text-to-video pre-trained model with 30Bational parameters and the ability to generate videos up to 204 frames in length.<n>A deep compression Vari Autoencoder, Video-VAE, is designed for video generation tasks, achieving 16x16 spatial and 8x temporal compression ratios.<n>Step-Video-T2V's performance is evaluated on a novel video generation benchmark, Step-Video-T2V-Eval, demonstrating its state-of-the-art text-to-video quality.
arXiv Detail & Related papers (2025-02-14T15:58:10Z) - Vchitect-2.0: Parallel Transformer for Scaling Up Video Diffusion Models [89.79067761383855]
Vchitect-2.0 is a parallel transformer architecture designed to scale up video diffusion models for large-scale text-to-video generation.<n>By introducing a novel Multimodal Diffusion Block, our approach achieves consistent alignment between text descriptions and generated video frames.<n>To overcome memory and computational bottlenecks, we propose a Memory-efficient Training framework.
arXiv Detail & Related papers (2025-01-14T21:53:11Z) - STIV: Scalable Text and Image Conditioned Video Generation [84.2574247093223]
We present a simple and scalable text-image-conditioned video generation method, named STIV.<n>Our framework integrates image condition into a Diffusion Transformer (DiT) through frame replacement, while incorporating text conditioning.<n> STIV can be easily extended to various applications, such as video prediction, frame, multi-view generation, and long video generation.
arXiv Detail & Related papers (2024-12-10T18:27:06Z) - xGen-VideoSyn-1: High-fidelity Text-to-Video Synthesis with Compressed Representations [120.52120919834988]
xGen-SynVideo-1 is a text-to-video (T2V) generation model capable of producing realistic scenes from textual descriptions.
VidVAE compresses video data both spatially and temporally, significantly reducing the length of visual tokens.
DiT model incorporates spatial and temporal self-attention layers, enabling robust generalization across different timeframes and aspect ratios.
arXiv Detail & Related papers (2024-08-22T17:55:22Z) - LAMP: Learn A Motion Pattern for Few-Shot-Based Video Generation [44.220329202024494]
We present a few-shot-based tuning framework, LAMP, which enables text-to-image diffusion model Learn A specific Motion Pattern with 816 videos on a single GPU.
Specifically, we design a first-frame-conditioned pipeline that uses an off-the-shelf text-to-image model for content generation.
To capture the features of temporal dimension, we expand the pretrained 2D convolution layers of the T2I model to our novel temporal-spatial motion learning layers.
arXiv Detail & Related papers (2023-10-16T19:03:19Z) - LAVIE: High-Quality Video Generation with Cascaded Latent Diffusion
Models [133.088893990272]
We learn a high-quality text-to-video (T2V) generative model by leveraging a pre-trained text-to-image (T2I) model as a basis.
We propose LaVie, an integrated video generation framework that operates on cascaded video latent diffusion models.
arXiv Detail & Related papers (2023-09-26T17:52:03Z) - Make-A-Video: Text-to-Video Generation without Text-Video Data [69.20996352229422]
Make-A-Video is an approach for translating the tremendous recent progress in Text-to-Image (T2I) generation to Text-to-Video (T2V)
We design a simple yet effective way to build on T2I models with novel and effective spatial-temporal modules.
In all aspects, spatial and temporal resolution, faithfulness to text, and quality, Make-A-Video sets the new state-of-the-art in text-to-video generation.
arXiv Detail & Related papers (2022-09-29T13:59:46Z)
This list is automatically generated from the titles and abstracts of the papers in this site.
This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.