NeurIPS 2024 Papers

We propose a novel method for learning representations of poses for 3D deformable objects, which specializes in 1) disentangling pose information from the object's identity, 2) facilitating the learning of pose variations, and 3) transferring pose information to other object identities. Based on these properties, our method enables the generation of 3D deformable objects with diversity in both identities and poses, using variations of a single object. It does not require explicit shape parameterization such as skeletons or joints, point-level or shape-level correspondence supervision, or variations of the target object for pose transfer. To achieve pose disentanglement, compactness for generative models, and transferability, we first design the pose extractor to represent the pose as a keypoint-based hybrid representation and the pose applier to learn an implicit deformation field. To better distill pose information from the object's geometry, we propose the implicit pose applier to output an intrinsic mesh property, the face Jacobian. Once the extracted pose information is transferred to the target object, the pose applier is fine-tuned in a self-supervised manner to better describe the target object's shapes with pose variations. The extracted poses are also used to train a cascaded diffusion model to enable the generation of novel poses. Our experiments with the DeformThings4D and Human datasets demonstrate state-of-the-art performance in pose transfer and the ability to generate diverse deformed shapes with various objects and poses.

The diffusion model, a new generative modeling paradigm, has achieved significant success in generating images, audio, video, and text. It has been adapted for sequence-to-sequence text generation (Seq2Seq) through DiffuSeq, termed the S2S-Diffusion model. Existing S2S-Diffusion models predominantly rely on fixed or hand-crafted rules to schedule noise during the diffusion and denoising processes. However, these models are limited by non-contextualized noise, which fails to fully consider the characteristics of Seq2Seq tasks. In this paper, we propose the Meta-Diffu$B$ framework—a novel scheduler-exploiter S2S-Diffusion paradigm designed to overcome the limitations of existing S2S-Diffusion models. We employ Meta-Exploration to train an additional scheduler model dedicated to scheduling contextualized noise for each sentence. Our exploiter model, an S2S-Diffusion model, leverages the noise scheduled by our scheduler model for updating and generation. Meta-Diffu$B$ achieves state-of-the-art performance compared to previous S2S-Diffusion models and fine-tuned pre-trained language models (PLMs) across four Seq2Seq benchmark datasets. We further investigate and visualize the impact of Meta-Diffu$B$'s noise scheduling on the generation of sentences with varying difficulties. Additionally, our scheduler model can function as a "plug-and-play" model to enhance DiffuSeq without the need for fine-tuning during the inference stage.

Binocular 3D human pose estimation (HPE), reconstructing a 3D pose from 2D poses of two views, offers practical advantages by combining multiview geometry with the convenience of a monocular setup. However, compared to a multiview setup, the reduction in the number of cameras increases uncertainty in 3D reconstruction. To address this issue, we leverage the diffusion model, which has shown success in monocular 3D HPE by recovering 3D poses from noisy data with high uncertainty. Yet, the uncertainty distribution of initial 3D poses remains unknown. Considering that 3D errors stem from 2D errors within geometric constraints, we recognize that the uncertainties of 3D and 2D are integrated in a binocular configuration, with the initial 2D uncertainty being well-defined. Based on this insight, we propose Dual-Diffusion specifically for Binocular 3D HPE, simultaneously denoising the uncertainties in 2D and 3D, and recovering plausible and accurate results. Additionally, we introduce Z-embedding as an additional condition for denoising and implement baseline-width-related pose normalization to enhance the model flexibility for various baseline settings. This is crucial as 3D error influence factors encompass depth and baseline width. Extensive experiments validate the effectiveness of our Dual-Diffusion in 2D refinement and 3D estimation. The code and models are available at https://github.com/sherrywan/Dual-Diffusion.

Temporal action segmentation and long-term action anticipation are two popular vision tasks for the temporal analysis of actions in videos. Despite apparent relevance and potential complementarity, these two problems have been investigated as separate and distinct tasks. In this work, we tackle these two problems, action segmentation, and action anticipation, jointly using a unified diffusion model dubbed ActFusion. The key idea to unification is to train the model to effectively handle both visible and invisible parts of the sequence in an integrated manner; the visible part is for temporal segmentation, and the invisible part is for future anticipation. To this end, we introduce a new anticipative masking strategy during training in which a late part of the video frames is masked as invisible, and learnable tokens replace these frames to learn to predict the invisible future. Experimental results demonstrate the bi-directional benefits between action segmentation and anticipation. ActFusion achieves the state-of-the-art performance across the standard benchmarks of 50 Salads, Breakfast, and GTEA, outperforming task-specific models in both of the two tasks with a single unified model through joint learning.

The recent introduction of Diffusion Transformers (DiTs) has demonstrated exceptional capabilities in image generation by using a different backbone architecture, departing from traditional U-Nets and embracing the scalable nature of transformers. Despite their advanced capabilities, the wide deployment of DiTs, particularly for real-time applications, is currently hampered by considerable computational demands at the inference stage. Post-training Quantization (PTQ) has emerged as a fast and data-efficient solution that can significantly reduce computation and memory footprint by using low-bit weights and activations. However, its applicability to DiTs has not yet been explored and faces non-trivial difficulties due to the unique design of DiTs. In this paper, we propose PTQ4DiT, a specifically designed PTQ method for DiTs. We discover two primary quantization challenges inherent in DiTs, notably the presence of salient channels with extreme magnitudes and the temporal variability in distributions of salient activation over multiple timesteps. To tackle these challenges, we propose Channel-wise Salience Balancing (CSB) and Spearmen's $\rho$-guided Salience Calibration (SSC). CSB leverages the complementarity property of channel magnitudes to redistribute the extremes, alleviating quantization errors for both activations and weights. SSC extends this approach by dynamically adjusting the balanced salience to capture the temporal variations in activation. Additionally, to eliminate extra computational costs caused by PTQ4DiT during inference, we design an offline re-parameterization strategy for DiTs. Experiments demonstrate that our PTQ4DiT successfully quantizes DiTs to 8-bit precision (W8A8) while preserving comparable generation ability and further enables effective quantization to 4-bit weight precision (W4A8) for the first time.

The most advanced diffusion models have recently adopted increasingly deep stacked networks (e.g., U-Net or Transformer) to promote the generative emergence capabilities of vision generation models similar to large language models (LLMs). However, progressively deeper stacked networks will intuitively cause numerical propagation errors and reduce noisy prediction capabilities on generative data, which hinders massively deep scalable training of vision generation models. In this paper, we first uncover the nature that neural networks being able to effectively perform generative denoising lies in the fact that the intrinsic residual unit has consistent dynamic property with the input signal's reverse diffusion process, thus supporting excellent generative abilities. Afterwards, we stand on the shoulders of two common types of deep stacked networks to propose a unified and massively scalable Neural Residual Diffusion Models framework (Neural-RDM for short), which is a simple yet meaningful change to the common architecture of deep generative networks by introducing a series of learnable gated residual parameters that conform to the generative dynamics. Experimental results on various generative tasks show that the proposed neural residual models obtain state-of-the-art scores on image's and video's generative benchmarks. Rigorous theoretical proofs and extensive experiments also demonstrate the advantages of this simple gated residual mechanism consistent with dynamic modeling in improving the fidelity and consistency of generated content and supporting large-scale scalable training.

Given an unconditional diffusion model and a predictor for a target property of interest (e.g., a classifier), the goal of training-free guidance is to generate samples with desirable target properties without additional training. Existing methods, though effective in various individual applications, often lack theoretical grounding and rigorous testing on extensive benchmarks. As a result, they could even fail on simple tasks, and applying them to a new problem becomes unavoidably difficult. This paper introduces a novel algorithmic framework encompassing existing methods as special cases, unifying the study of training-free guidance into the analysis of an algorithm-agnostic design space. Via theoretical and empirical investigation, we propose an efficient and effective hyper-parameter searching strategy that can be readily applied to any downstream task. We systematically benchmark across 7 diffusion models on 16 tasks with 40 targets, and improve performance by 8.5% on average. Our framework and benchmark offer a solid foundation for conditional generation in a training-free manner.

Foundation models have emerged as powerful tools across various domains including language, vision, and multimodal tasks. While prior works have addressed unsupervised semantic segmentation, they significantly lag behind supervised models. In this paper, we use a diffusion UNet encoder as a foundation vision encoder and introduce DiffCut, an unsupervised zero-shot segmentation method that solely harnesses the output features from the final self-attention block. Through extensive experimentation, we demonstrate that using these diffusion features in a graph based segmentation algorithm, significantly outperforms previous state-of-the-art methods on zero-shot segmentation. Specifically, we leverage a recursive Normalized Cut algorithm that regulates the granularity of detected objects and produces well-defined segmentation maps that precisely capture intricate image details. Our work highlights the remarkably accurate semantic knowledge embedded within diffusion UNet encoders that could then serve as foundation vision encoders for downstream tasks.

Each voxel in a diffusion MRI (dMRI) image contains a spherical signal corresponding to the direction and strength of water diffusion in the brain. This paper advances the analysis of such spatio-spherical data by developing convolutional network layers that are equivariant to the $\mathbf{E(3) \times SO(3)}$ group and account for the physical symmetries of dMRI including rotations, translations, and reflections of space alongside voxel-wise rotations. Further, neuronal fibers are typically antipodally symmetric, a fact we leverage to construct highly efficient spatio-*hemispherical* graph convolutions to accelerate the analysis of high-dimensional dMRI data. In the context of sparse spherical fiber deconvolution to recover white matter microstructure, our proposed equivariant network layers yield substantial performance and efficiency gains, leading to better and more practical resolution of crossing neuronal fibers and fiber tractography. These gains are experimentally consistent across both simulation and in vivo human datasets.

Text-to-image diffusion generative models can generate high quality images at the cost of tedious prompt engineering. Controllability can be improved by introducing layout conditioning, however existing methods lack layout editing ability and fine-grained control over object attributes. The concept of multi-layer generation holds great potential to address these limitations, however generating image instances concurrently to scene composition limits control over fine-grained object attributes, relative positioning in 3D space and scene manipulation abilities. In this work, we propose a novel multi-stage generation paradigm that is designed for fine-grained control, flexibility and interactivity. To ensure control over instance attributes, we devise a novel training paradigm to adapt a diffusion model to generate isolated scene components as RGBA images with transparency information. To build complex images, we employ these pre-generated instances and introduce a multi-layer composite generation process that smoothly assembles components in realistic scenes. Our experiments show that our RGBA diffusion model is capable of generating diverse and high quality instances with precise control over object attributes. Through multi-layer composition, we demonstrate that our approach allows to build and manipulate images from highly complex prompts with fine-grained control over object appearance and location, granting a higher degree of control than competing methods.

Transformer-based Diffusion Probabilistic Models (DPMs) have shown more potential than CNN-based DPMs, yet their extensive computational requirements hinder widespread practical applications. To reduce the computation budget of transformer-based DPMs, this work proposes the Efficient Diffusion Transformer (EDT) framework. This framework includes a lightweight-design diffusion model architecture, and a training-free Attention Modulation Matrix and its alternation arrangement in EDT inspired by human-like sketching. Additionally, we propose a token relation-enhanced masking training strategy tailored explicitly for EDT to augment its token relation learning capability. Our extensive experiments demonstrate the efficacy of EDT. The EDT framework reduces training and inference costs and surpasses existing transformer-based diffusion models in image synthesis performance, thereby achieving a significant overall enhancement. With lower FID, EDT-S, EDT-B, and EDT-XL attained speed-ups of 3.93x, 2.84x, and 1.92x respectively in the training phase, and 2.29x, 2.29x, and 2.22x respectively in inference, compared to the corresponding sizes of MDTv2. Our code is available at https://github.com/xinwangChen/EDT.

Diffusion Probabilistic Models (DPMs) have been recently utilized to deal with various blind image restoration (IR) tasks, where they have demonstrated outstanding performance in terms of perceptual quality. However, the task-specific nature of existing solutions and the excessive computational costs related to their training, make such models impractical and challenging to use for different IR tasks than those that were initially trained for. This hinders their wider adoption especially by those who lack access to powerful computational resources and vast amounts of training data. In this work we aim to address the above issues and enable the successful adoption of DPMs in practical IR-related applications. Towards this goal, we propose a modular diffusion probabilistic IR framework (DP-IR), which allows us to combine the performance benefits of existing pre-trained state-of-the-art IR networks and generative DPMs, while it requires only the additional training of a small module (0.7M params) related to the particular IR task of interest. Moreover, the architecture of our proposed framework allows us to employ a sampling strategy that leads to at least four times reduction of neural function evaluations without any performance loss, while it can also be combined with existing acceleration techniques (e.g. DDIM). We evaluate our model on four benchmarks for the tasks of burst JDD-SR, dynamic scene deblurring, and super-resolution. Our method outperforms existing approaches in terms of perceptual quality while retaining a competitive performance in relation to fidelity metrics.

Text-to-Image (T2I) models have made significant advancements in recent years, but they still struggle to accurately capture intricate details specified in complex compositional prompts. While fine-tuning T2I models with reward objectives has shown promise, it suffers from "reward hacking" and may not generalize well to unseen prompt distributions. In this work, we propose Reward-based Noise Optimization (ReNO), a novel approach that enhances T2I models at inference by optimizing the initial noise based on the signal from one or multiple human preference reward models. Remarkably, solving this optimization problem with gradient ascent for 50 iterations yields impressive results on four different one-step models across two competitive benchmarks, T2I-CompBench and GenEval. Within a computational budget of 20-50 seconds, ReNO-enhanced one-step models consistently surpass the performance of all current open-source Text-to-Image models. Extensive user studies demonstrate that our model is preferred nearly twice as often compared to the popular SDXL model and is on par with the proprietary Stable Diffusion 3 with 8B parameters. Moreover, given the same computational resources, a ReNO-optimized one-step model outperforms widely-used open-source models such as SDXL and PixArt-alpha, highlighting the efficiency and effectiveness of ReNO in enhancing T2I model performance at inference time.

Protein optimization is a fundamental biological task aimed at enhancing theperformance of proteins by modifying their sequences. Computational methodsprimarily rely on evolutionary information (EI) encoded by protein languagemodels (PLMs) to predict fitness landscape for optimization. However, thesemethods suffer from a few limitations. (1) Evolutionary processes involve thesimultaneous consideration of multiple functional properties, often overshadowingthe specific property of interest. (2) Measurements of these properties tend to betailored to experimental conditions, leading to reduced generalizability of trainedmodels to novel proteins. To address these limitations, we introduce DenoisingProtein Language Models (DePLM), a novel approach that refines the evolutionaryinformation embodied in PLMs for improved protein optimization. Specifically, weconceptualize EI as comprising both property-relevant and irrelevant information,with the latter acting as “noise” for the optimization task at hand. Our approachinvolves denoising this EI in PLMs through a diffusion process conducted in therank space of property values, thereby enhancing model generalization and ensuringdataset-agnostic learning. Extensive experimental results have demonstrated thatDePLM not only surpasses the state-of-the-art in mutation effect prediction butalso exhibits strong generalization capabilities for novel proteins.

We present Diffusion-KTO, a novel approach for aligning text-to-image diffusion models by formulating the alignment objective as the maximization of expected human utility. Unlike previous methods, Diffusion-KTO does not require collecting pairwise preference data nor training a complex reward model. Instead, our objective uses per-image binary feedback signals, e.g. likes or dislikes, to align the model with human preferences. After fine-tuning using Diffusion-KTO, text-to-image diffusion models exhibit improved performance compared to existing techniques, including supervised fine-tuning and Diffusion-DPO, both in terms of human judgment and automatic evaluation metrics such as PickScore and ImageReward. Overall, Diffusion-KTO unlocks the potential of leveraging readily available per-image binary preference signals and broadens the applicability of aligning text-to-image diffusion models with human preferences.

With high-dimensional state spaces, visual reinforcement learning (RL) faces significant challenges in exploitation and exploration, resulting in low sample efficiency and training stability. As a time-efficient diffusion model, although consistency models have been validated in online state-based RL, it is still an open question whether it can be extended to visual RL. In this paper, we investigate the impact of non-stationary distribution and the actor-critic framework on consistency policy in online RL, and find that consistency policy was unstable during the training, especially in visual RL with the high-dimensional state space. To this end, we suggest sample-based entropy regularization to stabilize the policy training, and propose a consistency policy with prioritized proximal experience regularization (CP3ER) to improve sample efficiency. CP3ER achieves new state-of-the-art (SOTA) performance in 21 tasks across DeepMind control suite and Meta-world. To our knowledge, CP3ER is the first method to apply diffusion/consistency models to visual RL and demonstrates the potential of consistency models in visual RL.

Fine-grained visual classification (FGVC) involves classifying closely related subcategories. This task is inherently difficult due to the subtle differences between classes and the high intra-class variance. Moreover, FGVC datasets are typically small and challenging to gather, thus highlighting a significant need for effective data augmentation. Recent advancements in text-to-image diffusion models have introduced new possibilities for data augmentation in image classification. While these models have been used to generate training data for classification tasks, their effectiveness in full-dataset training of FGVC models remains under-explored. Recent techniques that rely on text-to-image generation or Img2Img methods, such as SDEdit, often struggle to generate images that accurately represent the class while modifying them to a degree that significantly increases the dataset's diversity. To address these challenges, we present SaSPA: Structure and Subject Preserving Augmentation. Contrary to recent methods, our method does not use real images as guidance, thereby increasing generation flexibility and promoting greater diversity. To ensure accurate class representation, we employ conditioning mechanisms, specifically by conditioning on image edges and subject representation. We conduct extensive experiments and benchmark SaSPA against both traditional and generative data augmentation techniques. SaSPA consistently outperforms all established baselines across multiple settings, including full dataset training and contextual bias. Additionally, our results reveal interesting patterns in using synthetic data for FGVC models; for instance, we find a relationship between the amount of real data used and the optimal proportion of synthetic data.

We study the problem of estimating the body movements of a camera wearer from egocentric videos. Current methods for ego-body pose estimation rely on temporally dense sensor data, such as IMU measurements from spatially sparse body parts like the head and hands. However, we propose that even temporally sparse observations, such as hand poses captured intermittently from egocentric videos during natural or periodic hand movements, can effectively constrain overall body motion. Naively applying diffusion models to generate full-body pose from head pose and sparse hand pose leads to suboptimal results. To overcome this, we develop a two-stage approach that decomposes the problem into temporal completion and spatial completion. First, our method employs masked autoencoders to impute hand trajectories by leveraging the spatiotemporal correlations between the head pose sequence and intermittent hand poses, providing uncertainty estimates. Subsequently, we employ conditional diffusion models to generate plausible full-body motions based on these temporally dense trajectories of the head and hands, guided by the uncertainty estimates from the imputation. The effectiveness of our methods was rigorously tested and validated through comprehensive experiments conducted on various HMD setup with AMASS and Ego-Exo4D datasets. Project page: https://sgchi.github.io/dsposer

Image editing serves as a practical yet challenging task considering the diverse demands from users, where one of the hardest parts is to precisely describe how the edited image should look like. In this work, we present a new form of editing, termed imitative editing, to help users exercise their creativity more conveniently. Concretely, to edit an image region of interest, users are free to directly draw inspiration from some in-the-wild references (e.g., some relative pictures come across online), without having to cope with the fit between the reference and the source. Such a design requires the system to automatically figure out what to expect from the reference to perform the editing. For this purpose, we propose a generative training framework, dubbed MimicBrush, which randomly selects two frames from a video clip, masks some regions of one frame, and learns to recover the masked regions using the information from the other frame. That way, our model, developed from a diffusion prior, is able to capture the semantic correspondence between separate images in a self-supervised manner. We experimentally show the effectiveness of our method under various test cases as well as its superiority over existing alternatives. We also construct a benchmark to facilitate further research.