We present an efficient text-to-video generation framework based on latent diffusion models, termed MagicVideo. MagicVideo can generate smooth video clips that are concordant with the given text descriptions. Due to a novel and efficient 3D U-Net design and modeling video distributions in a low-dimensional space, MagicVideo can synthesize video clips with 256x256 spatial resolution on a single GPU card, which takes around 64x fewer computations than the Video Diffusion Models (VDM) in terms of FLOPs. In specific, unlike existing works that directly train video models in the RGB space, we use a pre-trained VAE to map video clips into a low-dimensional latent space and learn the distribution of videos' latent codes via a diffusion model. Besides, we introduce two new designs to adapt the U-Net denoiser trained on image tasks to video data: a frame-wise lightweight adaptor for the image-to-video distribution adjustment and a directed temporal attention module to capture temporal dependencies across frames. Thus, we can exploit the informative weights of convolution operators from a text-to-image model for accelerating video training. To ameliorate the pixel dithering in the generated videos, we also propose a novel VideoVAE auto-encoder for better RGB reconstruction. We conduct extensive experiments and demonstrate that MagicVideo can generate high-quality video clips with either realistic or imaginary content. Refer to \url{//magicvideo.github.io/#} for more examples.
相關內容
ACM/IEEE第23屆模型驅動工程語言和系統國際會議，是模型驅動軟件和系統工程的首要會議系列，由ACM-SIGSOFT和IEEE-TCSE支持組織。自1998年以來，模型涵蓋了建模的各個方面，從語言和方法到工具和應用程序。模特的參加者來自不同的背景，包括研究人員、學者、工程師和工業專業人士。MODELS 2019是一個論壇，參與者可以圍繞建模和模型驅動的軟件和系統交流前沿研究成果和創新實踐經驗。今年的版本將為建模社區提供進一步推進建模基礎的機會，并在網絡物理系統、嵌入式系統、社會技術系統、云計算、大數據、機器學習、安全、開源等新興領域提出建模的創新應用以及可持續性。
官網鏈接： ·
anchor
·
MoDELS
·
解碼
·
Learning
·
Unpaired image-to-image translation (UNIT) aims to map images between two visual domains without paired training data. However, given a UNIT model trained on certain domains, it is difficult for current methods to incorporate new domains because they often need to train the full model on both existing and new domains. To address this problem, we propose a new domain-scalable UNIT method, termed as latent space anchoring, which can be efficiently extended to new visual domains and does not need to fine-tune encoders and decoders of existing domains. Our method anchors images of different domains to the same latent space of frozen GANs by learning lightweight encoder and regressor models to reconstruct single-domain images. In the inference phase, the learned encoders and decoders of different domains can be arbitrarily combined to translate images between any two domains without fine-tuning. Experiments on various datasets show that the proposed method achieves superior performance on both standard and domain-scalable UNIT tasks in comparison with the state-of-the-art methods.
As an efficient alternative to conventional full finetuning, parameter-efficient finetuning (PEFT) is becoming the prevailing method to adapt pretrained language models. In PEFT, a lightweight module is learned on each dataset while the underlying pretrained language model remains unchanged, resulting in multiple compact modules representing diverse skills when applied to various domains and tasks. In this paper, we propose to compose these parameter-efficient modules through linear arithmetic operations in the weight space, thereby integrating different module capabilities. Specifically, we first define addition and negation operators for the module, and then further compose these two basic operators to perform flexible arithmetic. Our approach requires \emph{no additional training} and enables highly flexible module composition. We apply different arithmetic operations to compose the parameter-efficient modules for (1) distribution generalization, (2) multi-tasking, (3) unlearning, and (4) domain transfer. Additionally, we extend our approach to detoxify Alpaca-LoRA, the latest instruction-tuned large language model based on LLaMA. Empirical results demonstrate that our approach produces new and effective parameter-efficient modules that significantly outperform existing ones across all settings.
Despite the tremendous success in text-to-image generative models, localized text-to-image generation (that is, generating objects or features at specific locations in an image while maintaining a consistent overall generation) still requires either explicit training or substantial additional inference time. In this work, we show that localized generation can be achieved by simply controlling cross attention maps during inference. With no additional training, model architecture modification or inference time, our proposed cross attention control (CAC) provides new open-vocabulary localization abilities to standard text-to-image models. CAC also enhances models that are already trained for localized generation when deployed at inference time. Furthermore, to assess localized text-to-image generation performance automatically, we develop a standardized suite of evaluations using large pretrained recognition models. Our experiments show that CAC improves localized generation performance with various types of location information ranging from bounding boxes to semantic segmentation maps, and enhances the compositional capability of state-of-the-art text-to-image generative models.
Text-to-image diffusion models have advanced towards more controllable generation via supporting various image conditions (e.g., depth map) beyond text. However, these models are learned based on the premise of perfect alignment between the text and image conditions. If this alignment is not satisfied, the final output could be either dominated by one condition, or ambiguity may arise, failing to meet user expectations. To address this issue, we present a training-free approach called "Decompose and Realign'' to further improve the controllability of existing models when provided with partially aligned conditions. The ``Decompose'' phase separates conditions based on pair relationships, computing scores individually for each pair. This ensures that each pair no longer has conflicting conditions. The "Realign'' phase aligns these independently calculated scores via a cross-attention mechanism to avoid new conflicts when combing them back. Both qualitative and quantitative results demonstrate the effectiveness of our approach in handling unaligned conditions, which performs favorably against recent methods and more importantly adds flexibility to the controllable image generation process.
The unprecedented success of image reconstruction approaches based on deep neural networks has revolutionised both the processing and the analysis paradigms in several applied disciplines. In the field of digital humanities, the task of digital reconstruction of ancient frescoes is particularly challenging due to the scarce amount of available training data caused by ageing, wear, tear and retouching over time. To overcome these difficulties, we consider the Deep Image Prior (DIP) inpainting approach which computes appropriate reconstructions by relying on the progressive updating of an untrained convolutional neural network so as to match the reliable piece of information in the image at hand while promoting regularisation elsewhere. In comparison with state-of-the-art approaches (based on variational/PDEs and patch-based methods), DIP-based inpainting reduces artefacts and better adapts to contextual/non-local information, thus providing a valuable and effective tool for art historians. As a case study, we apply such approach to reconstruct missing image contents in a dataset of highly damaged digital images of medieval paintings located into several chapels in the Mediterranean Alpine Arc and provide a detailed description on how visible and invisible (e.g., infrared) information can be integrated for identifying and reconstructing damaged image regions.
Generative Adversarial Networks (GANs) are powerful models able to synthesize data samples closely resembling the distribution of real data, yet the diversity of those generated samples is limited due to the so-called mode collapse phenomenon observed in GANs. Especially prone to mode collapse are conditional GANs, which tend to ignore the input noise vector and focus on the conditional information. Recent methods proposed to mitigate this limitation increase the diversity of generated samples, yet they reduce the performance of the models when similarity of samples is required. To address this shortcoming, we propose a novel method to selectively increase the diversity of GAN-generated samples. By adding a simple, yet effective regularization to the training loss function we encourage the generator to discover new data modes for inputs related to diverse outputs while generating consistent samples for the remaining ones. More precisely, we maximise the ratio of distances between generated images and input latent vectors scaling the effect according to the diversity of samples for a given conditional input. We show the superiority of our method in a synthetic benchmark as well as a real-life scenario of simulating data from the Zero Degree Calorimeter of ALICE experiment in LHC, CERN.
Foley sound generation aims to synthesise the background sound for multimedia content. Previous models usually employ a large development set with labels as input (e.g., single numbers or one-hot vector). In this work, we propose a diffusion model based system for Foley sound generation with text conditions. To alleviate the data scarcity issue, our model is initially pre-trained with large-scale datasets and fine-tuned to this task via transfer learning using the contrastive language-audio pertaining (CLAP) technique. We have observed that the feature embedding extracted by the text encoder can significantly affect the performance of the generation model. Hence, we introduce a trainable layer after the encoder to improve the text embedding produced by the encoder. In addition, we further refine the generated waveform by generating multiple candidate audio clips simultaneously and selecting the best one, which is determined in terms of the similarity score between the embedding of the candidate clips and the embedding of the target text label. Using the proposed method, our system ranks ${1}^{st}$ among the systems submitted to DCASE Challenge 2023 Task 7. The results of the ablation studies illustrate that the proposed techniques significantly improve sound generation performance. The codes for implementing the proposed system are available online.
3D content manipulation is an important computer vision task with many real-world applications (e.g., product design, cartoon generation, and 3D Avatar editing). Recently proposed 3D GANs can generate diverse photorealistic 3D-aware contents using Neural Radiance fields (NeRF). However, manipulation of NeRF still remains a challenging problem since the visual quality tends to degrade after manipulation and suboptimal control handles such as 2D semantic maps are used for manipulations. While text-guided manipulations have shown potential in 3D editing, such approaches often lack locality. To overcome these problems, we propose Local Editing NeRF (LENeRF), which only requires text inputs for fine-grained and localized manipulation. Specifically, we present three add-on modules of LENeRF, the Latent Residual Mapper, the Attention Field Network, and the Deformation Network, which are jointly used for local manipulations of 3D features by estimating a 3D attention field. The 3D attention field is learned in an unsupervised way, by distilling the zero-shot mask generation capability of CLIP to the 3D space with multi-view guidance. We conduct diverse experiments and thorough evaluations both quantitatively and qualitatively.
Conventionally, spatiotemporal modeling network and its complexity are the two most concentrated research topics in video action recognition. Existing state-of-the-art methods have achieved excellent accuracy regardless of the complexity meanwhile efficient spatiotemporal modeling solutions are slightly inferior in performance. In this paper, we attempt to acquire both efficiency and effectiveness simultaneously. First of all, besides traditionally treating H x W x T video frames as space-time signal (viewing from the Height-Width spatial plane), we propose to also model video from the other two Height-Time and Width-Time planes, to capture the dynamics of video thoroughly. Secondly, our model is designed based on 2D CNN backbones and model complexity is well kept in mind by design. Specifically, we introduce a novel multi-view fusion (MVF) module to exploit video dynamics using separable convolution for efficiency. It is a plug-and-play module and can be inserted into off-the-shelf 2D CNNs to form a simple yet effective model called MVFNet. Moreover, MVFNet can be thought of as a generalized video modeling framework and it can specialize to be existing methods such as C2D, SlowOnly, and TSM under different settings. Extensive experiments are conducted on popular benchmarks (i.e., Something-Something V1 & V2, Kinetics, UCF-101, and HMDB-51) to show its superiority. The proposed MVFNet can achieve state-of-the-art performance with 2D CNN's complexity.
High spectral dimensionality and the shortage of annotations make hyperspectral image (HSI) classification a challenging problem. Recent studies suggest that convolutional neural networks can learn discriminative spatial features, which play a paramount role in HSI interpretation. However, most of these methods ignore the distinctive spectral-spatial characteristic of hyperspectral data. In addition, a large amount of unlabeled data remains an unexploited gold mine for efficient data use. Therefore, we proposed an integration of generative adversarial networks (GANs) and probabilistic graphical models for HSI classification. Specifically, we used a spectral-spatial generator and a discriminator to identify land cover categories of hyperspectral cubes. Moreover, to take advantage of a large amount of unlabeled data, we adopted a conditional random field to refine the preliminary classification results generated by GANs. Experimental results obtained using two commonly studied datasets demonstrate that the proposed framework achieved encouraging classification accuracy using a small number of data for training.
小貼士
登錄享
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191