CVPR 2023 Papers

While some studies have proven that Swin Transformer (Swin) with window self-attention (WSA) is suitable for single image super-resolution (SR), the plain WSA ignores the broad regions when reconstructing high-resolution images due to a limited receptive field. In addition, many deep learning SR methods suffer from intensive computations. To address these problems, we introduce the N-Gram context to the low-level vision with Transformers for the first time. We define N-Gram as neighboring local windows in Swin, which differs from text analysis that views N-Gram as consecutive characters or words. N-Grams interact with each other by sliding-WSA, expanding the regions seen to restore degraded pixels. Using the N-Gram context, we propose NGswin, an efficient SR network with SCDP bottleneck taking multi-scale outputs of the hierarchical encoder. Experimental results show that NGswin achieves competitive performance while maintaining an efficient structure when compared with previous leading methods. Moreover, we also improve other Swin-based SR methods with the N-Gram context, thereby building an enhanced model: SwinIR-NG. Our improved SwinIR-NG outperforms the current best lightweight SR approaches and establishes state-of-the-art results. Codes are available at https://github.com/rami0205/NGramSwin.

We develop a simple yet surprisingly effective implicit representing scheme called Multiplicative Fourier Level of Detail (MFLOD) motivated by the recent success of multiplicative filter network. Built on multi-resolution feature grid/volume (e.g., the sparse voxel octree), each level's feature is first modulated by a sinusoidal function and then element-wisely multiplied by a linear transformation of previous layer's representation in a layer-to-layer recursive manner, yielding the scale-aggregated encodings for a subsequent simple linear forward to get final output. In contrast to previous hybrid representations relying on interleaved multilevel fusion and nonlinear activation-based decoding, MFLOD could be elegantly characterized as a linear combination of sine basis functions with varying amplitude, frequency, and phase upon the learned multilevel features, thus offering great feasibility in Fourier analysis. Comprehensive experimental results on implicit neural representation learning tasks including image fitting, 3D shape representation, and neural radiance fields well demonstrate the superior quality and generalizability achieved by the proposed MFLOD scheme.

Image-text retrieval is a fundamental task to bridge vision and language by exploiting various strategies to fine-grained alignment between regions and words. This is still tough mainly because of one-to-many correspondence, where a set of matches from another modality can be accessed by a random query. While existing solutions to this problem including multi-point mapping, probabilistic distribution, and geometric embedding have made promising progress, one-to-many correspondence is still under-explored. In this work, we develop a Multilateral Semantic Relations Modeling (termed MSRM) for image-text retrieval to capture the one-to-many correspondence between multiple samples and a given query via hypergraph modeling. Specifically, a given query is first mapped as a probabilistic embedding to learn its true semantic distribution based on Mahalanobis distance. Then each candidate instance in a mini-batch is regarded as a hypergraph node with its mean semantics while a Gaussian query is modeled as a hyperedge to capture the semantic correlations beyond the pair between candidate points and the query. Comprehensive experimental results on two widely used datasets demonstrate that our MSRM method can outperform state-of-the-art methods in the settlement of multiple matches while still maintaining the comparable performance of instance-level matching. Our codes and checkpoints will be released soon.

By optimizing the rate-distortion-realism trade-off, generative compression approaches produce detailed, realistic images, even at low bit rates, instead of the blurry reconstructions produced by rate-distortion optimized models. However, previous methods do not explicitly control how much detail is synthesized, which results in a common criticism of these methods: users might be worried that a misleading reconstruction far from the input image is generated. In this work, we alleviate these concerns by training a decoder that can bridge the two regimes and navigate the distortion-realism trade-off. From a single compressed representation, the receiver can decide to either reconstruct a low mean squared error reconstruction that is close to the input, a realistic reconstruction with high perceptual quality, or anything in between. With our method, we set a new state-of-the-art in distortion-realism, pushing the frontier of achievable distortion-realism pairs, i.e., our method achieves better distortions at high realism and better realism at low distortion than ever before.

While generative models produce high-quality images of concepts learned from a large-scale database, a user often wishes to synthesize instantiations of their own concepts (for example, their family, pets, or items). Can we teach a model to quickly acquire a new concept, given a few examples? Furthermore, can we compose multiple new concepts together? We propose Custom Diffusion, an efficient method for augmenting existing text-to-image models. We find that only optimizing a few parameters in the text-to-image conditioning mechanism is sufficiently powerful to represent new concepts while enabling fast tuning ( 6 minutes). Additionally, we can jointly train for multiple concepts or combine multiple fine-tuned models into one via closed-form constrained optimization. Our fine-tuned model generates variations of multiple new concepts and seamlessly composes them with existing concepts in novel settings. Our method outperforms or performs on par with several baselines and concurrent works in both qualitative and quantitative evaluations, while being memory and computationally efficient.

Image captioning models are known to perpetuate and amplify harmful societal bias in the training set. In this work, we aim to mitigate such gender bias in image captioning models. While prior work has addressed this problem by forcing models to focus on people to reduce gender misclassification, it conversely generates gender-stereotypical words at the expense of predicting the correct gender. From this observation, we hypothesize that there are two types of gender bias affecting image captioning models: 1) bias that exploits context to predict gender, and 2) bias in the probability of generating certain (often stereotypical) words because of gender. To mitigate both types of gender biases, we propose a framework, called LIBRA, that learns from synthetically biased samples to decrease both types of biases, correcting gender misclassification and changing gender-stereotypical words to more neutral ones.

Video generation remains a challenging task due to spatiotemporal complexity and the requirement of synthesizing diverse motions with temporal consistency. Previous works attempt to generate videos in arbitrary lengths either in an autoregressive manner or regarding time as a continuous signal. However, they struggle to synthesize detailed and diverse motions with temporal coherence and tend to generate repetitive scenes after a few time steps. In this work, we argue that a single time-agnostic latent vector of style-based generator is insufficient to model various and temporally-consistent motions. Hence, we introduce additional time-dependent motion styles to model diverse motion patterns. In addition, a Motion Style Attention modulation mechanism, dubbed as MoStAtt, is proposed to augment frames with vivid dynamics for each specific scale (i.e., layer), which assigns attention score for each motion style w.r.t deconvolution filter weights in the target synthesis layer and softly attends different motion styles for weight modulation. Experimental results show our model achieves state-of-the-art performance on four unconditional 256^2 video synthesis benchmarks trained with only 3 frames per clip and produces better qualitative results with respect to dynamic motions. Code and videos have been made available at https://github.com/xiaoqian-shen/MoStGAN-V.

In this work, we propose an ID-preserving talking head generation framework, which advances previous methods in two aspects. First, as opposed to interpolating from sparse flow, we claim that dense landmarks are crucial to achieving accurate geometry-aware flow fields. Second, inspired by face-swapping methods, we adaptively fuse the source identity during synthesis, so that the network better preserves the key characteristics of the image portrait. Although the proposed model surpasses prior generation fidelity on established benchmarks, personalized fine-tuning is still needed to further make the talking head generation qualified for real usage. However, this process is rather computationally demanding that is unaffordable to standard users. To alleviate this, we propose a fast adaptation model using a meta-learning approach. The learned model can be adapted to a high-quality personalized model as fast as 30 seconds. Last but not least, a spatial-temporal enhancement module is proposed to improve the fine details while ensuring temporal coherency. Extensive experiments prove the significant superiority of our approach over the state of the arts in both one-shot and personalized settings.

Existing methods proposed for hand reconstruction tasks usually parameterize a generic 3D hand model or predict hand mesh positions directly. The parametric representations consisting of hand shapes and rotational poses are more stable, while the non-parametric methods can predict more accurate mesh positions. In this paper, we propose to reconstruct meshes and estimate MANO parameters of two hands from a single RGB image simultaneously to utilize the merits of two kinds of hand representations. To fulfill this target, we propose novel Mesh-Mano interaction blocks (MMIBs), which take mesh vertices positions and MANO parameters as two kinds of query tokens. MMIB consists of one graph residual block to aggregate local information and two transformer encoders to model long-range dependencies. The transformer encoders are equipped with different asymmetric attention masks to model the intra-hand and inter-hand attention, respectively. Moreover, we introduce the mesh alignment refinement module to further enhance the mesh-image alignment. Extensive experiments on the InterHand2.6M benchmark demonstrate promising results over the state-of-the-art hand reconstruction methods.

We present a novel framework for exemplar based image translation. Recent advanced methods for this task mainly focus on establishing cross-domain semantic correspondence, which sequentially dominates image generation in the manner of local style control. Unfortunately, cross domain semantic matching is challenging; and matching errors ultimately degrade the quality of generated images. To overcome this challenge, we improve the accuracy of matching on the one hand, and diminish the role of matching in image generation on the other hand. To achieve the former, we propose a masked and adaptive transformer (MAT) for learning accurate cross-domain correspondence, and executing context-aware feature augmentation. To achieve the latter, we use source features of the input and global style codes of the exemplar, as supplementary information, for decoding an image. Besides, we devise a novel contrastive style learning method, for acquire quality-discriminative style representations, which in turn benefit high-quality image generation. Experimental results show that our method, dubbed MATEBIT, performs considerably better than state-of-the-art methods, in diverse image translation tasks.

When capturing and storing images, devices inevitably introduce noise. Reducing this noise is a critical task called image denoising. Deep learning has become the de facto method for image denoising, especially with the emergence of Transformer-based models that have achieved notable state-of-the-art results on various image tasks. However, deep learning-based methods often suffer from a lack of generalization ability. For example, deep models trained on Gaussian noise may perform poorly when tested on other noise distributions. To address this issue, we present a novel approach to enhance the generalization performance of denoising networks, known as masked training. Our method involves masking random pixels of the input image and reconstructing the missing information during training. We also mask out the features in the self-attention layers to avoid the impact of training-testing inconsistency. Our approach exhibits better generalization ability than other deep learning models and is directly applicable to real-world scenarios. Additionally, our interpretability analysis demonstrates the superiority of our method.

Masked Image Modeling (MIM) achieves outstanding success in self-supervised representation learning. Unfortunately, MIM models typically have huge computational burden and slow learning process, which is an inevitable obstacle for their industrial applications. Although the lower layers play the key role in MIM, existing MIM models conduct reconstruction task only at the top layer of encoder. The lower layers are not explicitly guided and the interaction among their patches is only used for calculating new activations. Considering the reconstruction task requires non-trivial inter-patch interactions to reason target signals, we apply it to multiple local layers including lower and upper layers. Further, since the multiple layers expect to learn the information of different scales, we design local multi-scale reconstruction, where the lower and upper layers reconstruct fine-scale and coarse-scale supervision signals respectively. This design not only accelerates the representation learning process by explicitly guiding multiple layers, but also facilitates multi-scale semantical understanding to the input. Extensive experiments show that with significantly less pre-training burden, our model achieves comparable or better performance on classification, detection and segmentation tasks than existing MIM models.

Self supervision and natural language supervision have emerged as two exciting ways to train general purpose image encoders which excel at a variety of downstream tasks. Recent works such as M3AE (Geng et al 2022) and SLIP (Mu et al 2022) have suggested that these approaches can be effectively combined, but most notably their results use small (<20M examples) pre-training datasets and don't effectively reflect the large-scale regime (>100M samples) that is commonly used for these approaches. Here we investigate whether a similar approach can be effective when trained with a much larger amount of data. We find that a combination of two state of the art approaches: masked auto-encoders, MAE (He et al 2021) and contrastive language image pre-training, CLIP (Radford et al 2021) provides a benefit over CLIP when trained on a corpus of 11.3M image-text pairs, but little to no benefit (as evaluated on a suite of common vision tasks) over CLIP when trained on a large corpus of 1.4B images. Our work provides some much needed clarity into the effectiveness (or lack thereof) of self supervision for large-scale image-text training.

Recent conditional image generation methods produce images of remarkable diversity, fidelity and realism. However, the majority of these methods allow conditioning only on labels or text prompts, which limits their level of control over the generation result. In this paper, we introduce MaskSketch, an image generation method that allows spatial conditioning of the generation result using a guiding sketch as an extra conditioning signal during sampling. MaskSketch utilizes a pre-trained masked generative transformer, requiring no model training or paired supervision, and works with input sketches of different levels of abstraction. We show that intermediate self-attention maps of a masked generative transformer encode important structural information of the input image, such as scene layout and object shape, and we propose a novel sampling method based on this observation to enable structure-guided generation. Our results show that MaskSketch achieves high image realism and fidelity to the guiding structure. Evaluated on standard benchmark datasets, MaskSketch outperforms state-of-the-art methods for sketch-to-image translation, as well as unpaired image-to-image translation approaches. The code can be found on our project website: https://masksketch.github.io/

This paper presents a simple yet effective framework MaskCLIP, which incorporates a newly proposed masked self-distillation into contrastive language-image pretraining. The core idea of masked self-distillation is to distill representation from a full image to the representation predicted from a masked image. Such incorporation enjoys two vital benefits. First, masked self-distillation targets local patch representation learning, which is complementary to vision-language contrastive focusing on text-related representation. Second, masked self-distillation is also consistent with vision-language contrastive from the perspective of training objective as both utilize the visual encoder for feature aligning, and thus is able to learn local semantics getting indirect supervision from the language. We provide specially designed experiments with a comprehensive analysis to validate the two benefits. Symmetrically, we also introduce the local semantic supervision into the text branch, which further improves the pretraining performance. With extensive experiments, we show that MaskCLIP, when applied to various challenging downstream tasks, achieves superior results in linear probing, finetuning, and zero-shot performance with the guidance of the language encoder. We will release the code and data after the publication.

There has been a recent explosion of impressive generative models that can produce high quality images (or videos) conditioned on text descriptions. However, all such approaches rely on conditional sentences that contain unambiguous descriptions of scenes and main actors in them. Therefore employing such models for more complex task of story visualization, where naturally references and co-references exist, and one requires to reason about when to maintain consistency of actors and backgrounds across frames/scenes, and when not to, based on story progression, remains a challenge. In this work, we address the aforementioned challenges and propose a novel autoregressive diffusion-based framework with a visual memory module that implicitly captures the actor and background context across the generated frames. Sentence-conditioned soft attention over the memories enables effective reference resolution and learns to maintain scene and actor consistency when needed. To validate the effectiveness of our approach, we extend the MUGEN dataset and introduce additional characters, backgrounds and referencing in multi-sentence storylines. Our experiments for story generation on the MUGEN, the PororoSV and the FlintstonesSV dataset show that our method not only outperforms prior state-of-the-art in generating frames with high visual quality, which are consistent with the story, but also models appropriate correspondences between the characters and the background.

Recently, DreamFusion demonstrated the utility of a pretrained text-to-image diffusion model to optimize Neural Radiance Fields (NeRF), achieving remarkable text-to-3D synthesis results. However, the method has two inherent limitations: 1) optimization of the NeRF representation is extremely slow, 2) NeRF is supervised by images at a low resolution (64x64), thus leading to low-quality 3D models with a long wait time. In this paper, we address these limitations by utilizing a two-stage coarse-to-fine optimization framework. In the first stage, we use a sparse 3D neural representation to accelerate optimization while using a low-resolution diffusion prior. In the second stage, we use a textured mesh model initialized from the coarse neural representation, allowing us to perform optimization with a very efficient differentiable renderer interacting with high-resolution images. Our method, dubbed Magic3D, can create a 3D mesh model in 40 minutes, 2x faster than DreamFusion (reportedly taking 1.5 hours on average), while achieving 8x higher resolution. User studies show 61.7% raters to prefer our approach than DreamFusion. Together with the image-conditioned generation capabilities, we provide users with new ways to control 3D synthesis, opening up new avenues to various creative applications.

Advancements in the generation quality of various Generative Models (GMs) has made it necessary to not only perform binary manipulation detection but also localize the modified pixels in an image. However, prior works termed as passive for manipulation localization exhibit poor generalization performance over unseen GMs and attribute modifications. To combat this issue, we propose a proactive scheme for manipulation localization, termed MaLP. We encrypt the real images by adding a learned template. If the image is manipulated by any GM, this added protection from the template not only aids binary detection but also helps in identifying the pixels modified by the GM. The template is learned by leveraging local and global-level features estimated by a two-branch architecture. We show that MaLP performs better than prior passive works. We also show the generalizability of MaLP by testing on 22 different GMs, providing a benchmark for future research on manipulation localization. Finally, we show that MaLP can be used as a discriminator for improving the generation quality of GMs. Our models/codes are available at www.github.com/vishal3477/pro_loc.