Computer Vision – ECCV 2024

Masked Image Modeling (MIM) has emerged as a promising method for deriving visual representations from unlabeled image data by predicting missing pixels from masked portions of images. It excels in region-aware learning and provides strong initializations for various tasks, but struggles to capture high-level semantics without further supervised fine-tuning, likely due to the low-level nature of its pixel reconstruction objective. A promising yet unrealized framework is learning representations through masked reconstruction in latent space, combining the locality of MIM with the high-level targets. However, this approach poses significant training challenges as the reconstruction targets are learned in conjunction with the model, potentially leading to trivial or suboptimal solutions. Our study is among the first to thoroughly analyze and address the challenges of such framework, which we refer to as Latent MIM. Through a series of carefully designed experiments and extensive analysis, we identify the source of these challenges, including representation collapsing for joint online/target optimization, learning objectives, the high region correlation in latent space and decoding conditioning. By sequentially addressing these issues, we demonstrate that latent MIM can indeed learn high-level representations while retaining the benefits of MIM models. Code is available at https://​github.​com/​yibingwei-1/​LatentMIM.

Animating stylized characters to match a reference motion sequence is a highly demanded task in film and gaming industries. Existing methods mostly focus on rigid deformations of characters’ body, neglecting local deformations on the apparel driven by physical dynamics. They deform apparel the same way as the body, leading to results with limited details and unrealistic artifacts, e.g. body-apparel penetration. In contrast, we present a novel method aiming for high-quality motion transfer with realistic apparel animation. As existing datasets lack annotations necessary for generating realistic apparel animations, we build a new dataset named MMDMC, which combines stylized characters from the MikuMikuDance community with real-world Motion Capture data. We then propose a data-driven pipeline that learns to disentangle body and apparel deformations via two neural deformation modules. For body parts, we propose a geodesic attention block to effectively incorporate semantic priors into skeletal body deformation to tackle complex body shapes for stylized characters. Since apparel motion can significantly deviate from respective body joints, we propose to model apparel deformation in a non-linear vertex displacement field conditioned on its historic states. Extensive experiments show that our method produces results with superior quality for various types of apparel. Our dataset is released in https://​github.​com/​rongakowang/​MMDMC.

Inspired by cognitive theories, we introduce AnyHome, a framework that translates any text into well-structured and textured indoor scenes at a house-scale. By prompting Large Language Models (LLMs) with designed templates, our approach converts provided textual narratives into amodal structured representations. These representations guarantee consistent and realistic spatial layouts by directing the synthesis of a geometry mesh within defined constraints. A Score Distillation Sampling process is then employed to refine the geometry, followed by an egocentric inpainting process that adds lifelike textures to it. AnyHome stands out with its editability, customizability, diversity, and realism. The structured representations for scenes allow for extensive editing at varying levels of granularity. Capable of interpreting texts ranging from simple labels to detailed narratives, AnyHome generates detailed geometries and textures that outperform existing methods in both quantitative and qualitative measures.

We propose the SAL (Segment Anything in Lidar) method consisting of a text-promptable zero-shot model for segmenting and classifying any object in Lidar, and a pseudo-labeling engine that facilitates model training without manual supervision. While the established paradigm for Lidar Panoptic Segmentation (LPS) relies on manual supervision for a handful of object classes defined a priori, we utilize 2D vision foundation models to generate 3D supervision “for free”. Our pseudo-labels consist of instance masks and corresponding CLIP tokens, which we lift to Lidar using calibrated multi-modal data. By training our model on these labels, we distill the 2D foundation models into our Lidar SAL model. Even without manual labels, our model achieves \(91\%\) in terms of class-agnostic segmentation and \(54\%\) in terms of zero-shot LPS of the fully supervised state-of-the-art. Furthermore, we outperform several baselines that do not distill but only lift image features to 3D. More importantly, we demonstrate that SAL supports arbitrary class prompts, can be easily extended to new datasets, and shows significant potential to improve with increasing amounts of self-labeled data. We release all models and the code.

Large, pretrained latent diffusion models (LDMs) have demonstrated an extraordinary ability to generate creative content, specialize to user data through few-shot fine-tuning, and condition their output on other modalities, such as semantic maps. However, are they usable as large-scale data generators, e.g., to improve tasks in the perception stack, like semantic segmentation? We investigate this question in the context of autonomous driving, and answer it with a resounding “yes”. We propose an efficient data generation pipeline termed DGInStyle. First, we examine the problem of specializing a pretrained LDM to semantically-controlled generation within a narrow domain. Second, we propose a Style Swap technique to endow the rich generative prior with the learned semantic control. Third, we design a Multi-resolution Latent Fusion technique to overcome the bias of LDMs towards dominant objects. Using DGInStyle, we generate a diverse dataset of street scenes, train a domain-agnostic semantic segmentation model on it, and evaluate the model on multiple popular autonomous driving datasets. Our approach consistently increases the performance of several domain generalization methods compared to the previous state-of-the-art methods. The source code and the generated dataset are available at dginstyle.​github.​io.

We present a 3D modeling method which enables end-users to refine or detailize 3D shapes using machine learning, expanding the capabilities of AI-assisted 3D content creation. Given a coarse voxel shape (e.g., one produced with a simple box extrusion tool or via generative modeling), a user can directly “paint” desired target styles representing compelling geometric details, from input exemplar shapes, over different regions of the coarse shape. These regions are then up-sampled into high-resolution geometries which adhere with the painted styles. To achieve such controllable and localized 3D detailization, we build on top of a Pyramid GAN by making it masking-aware. We devise novel structural losses and priors to ensure that our method preserves both desired coarse structures and fine-grained features even if the painted styles are borrowed from diverse sources, e.g., different semantic parts and even different shape categories. Through extensive experiments, we show that our ability to localize details enables novel interactive creative workflows and applications. Our experiments further demonstrate that in comparison to prior techniques built on global detailization, our method generates structure-preserving, high-resolution stylized geometries with more coherent shape details and style transitions.

In this paper, we address the issue of scene-aware 3D human motion forecasting. A key challenge in this task is to predict future human motions that are coherent with the scene by modeling human-scene interactions. While recent works have demonstrated that explicit constraints on human-scene interactions can prevent the occurrence of ghost motion, they only provide constraints on partial human motion e.g., the global motion of the human or a few joints contacting the scene, leaving the rest of unconstrained. To address this limitation, we propose to represent the human-scene interaction using the mutual distance between the human body and the scene. Such mutual distances constrain both the local and global human motion, resulting in a whole-body motion constrained prediction. In particular, mutual distance constraints consist of two components, the signed distance of each vertex on the human mesh to the scene surface and the distance of basis scene points to the human mesh. We further introduce a global scene representation learned from a signed distance function (SDF) volume to ensure coherence between the global scene representation and the explicit constraint from the mutual distance. We develop a pipeline with two sequential steps: predicting the future mutual distances first, followed by forecasting future human motion. We explicitly ensure consistency between predicted poses and mutual distances during training. Extensive testing on both synthetic and real datasets demonstrates that our method consistently surpasses the performance of current state-of-the-art techniques.

Novel view synthesis from limited observations remains a crucial and ongoing challenge. In the realm of NeRF-based few-shot view synthesis, there is often a trade-off between the accuracy of the synthesized view and the efficiency of the 3D representation. To tackle this dilemma, we introduce a Few-Shot view synthesis framework based on 3D Gaussian Splatting, which facilitates real-time, photo-realistic synthesis from a minimal number of training views. FSGS employs an innovative Proximity-guided Gaussian Unpooling, specifically designed for sparse-view settings, to bridge the gap presented by the sparse initial point sets. This method involves the strategic placement of new Gaussians between existing ones, guided by a Gaussian proximity score, enhancing the adaptive density control. We have identified that Gaussian optimization can sometimes result in overly smooth textures and a propensity for overfitting when training views are limited. To mitigate these issues, FSGS introduces the synthesis of virtual views to replicate the parallax effect experienced during training, coupled with geometric regularization applied across both actual training and synthesized viewpoints. This strategy ensures that new Gaussians are placed in the most representative locations, fostering more accurate and detailed scene reconstruction. Our comprehensive evaluation across various datasets-including NeRF-Synthetic, LLFF, Shiny, and Mip-NeRF360 datasets-illustrates that FSGS not only delivers exceptional rendering quality but also achieves an inference speed more than 2000 times faster than existing state-of-the-art methods for sparse-view synthesis. Project webpage: https://​zehaozhu.​github.​io/​FSGS/​.

Panoramic images, capturing a 360\(^\circ \) field of view (FoV), encompass omnidirectional spatial information crucial for scene understanding. However, it is not only costly to obtain training-sufficient dense-annotated panoramas but also application-restricted when training models in a close-vocabulary setting. To tackle this problem, in this work, we define a new task termed Open Panoramic Segmentation (OPS), where models are trained with FoV-restricted pinhole images in the source domain in an open-vocabulary setting while evaluated with FoV-open panoramic images in the target domain, enabling the zero-shot open panoramic semantic segmentation ability of models. Moreover, we propose a model named OOOPS with a Deformable Adapter Network (DAN), which significantly improves zero-shot panoramic semantic segmentation performance. To further enhance the distortion-aware modeling ability from the pinhole source domain, we propose a novel data augmentation method called Random Equirectangular Projection (RERP) which is specifically designed to address object deformations in advance. Surpassing other state-of-the-art open-vocabulary semantic segmentation approaches, a remarkable performance boost on three panoramic datasets, WildPASS, Stanford2D3D, and Matterport3D, proves the effectiveness of our proposed OOOPS model with RERP on the OPS task, especially \({\textbf {+2.2}}\boldsymbol{\%}\) on outdoor WildPASS and \({\textbf {+2.4}}\boldsymbol{\%}\) mIoU on indoor Stanford2D3D. The source code is publicly available at OPS.

Learning feature correspondence is a foundational task in computer vision, holding immense importance for downstream applications such as visual odometry and 3D reconstruction. Despite recent progress in data-driven models, feature correspondence learning is still limited by the lack of accurate per-pixel correspondence labels. To overcome this difficulty, we introduce a new self-supervised scheme, imperative learning (IL), for training feature correspondence. It enables correspondence learning on arbitrary uninterrupted videos without any camera pose or depth labels, heralding a new era for self-supervised correspondence learning. Specifically, we formulated the problem of correspondence learning as a bilevel optimization, which takes the reprojection error from bundle adjustment as a supervisory signal for the model. It leads to a mutual improvement between the matching model and the bundle adjustment. To avoid large memory and computation overhead, we leverage the stationary point to efficiently back-propagate the implicit gradients through bundle adjustment. Through extensive experiments, we demonstrate superior performance on tasks including feature matching and pose estimation, in which we obtained an average of 30% accuracy gain over the state-of-the-art matching models.

We present a novel framework to reconstruct complete 3D human shapes from a given target image by leveraging monocular unconstrained images. The objective of this work is to reproduce high-quality details in regions of the reconstructed human body that are not visible in the input target. The proposed methodology addresses the limitations of existing approaches for reconstructing 3D human shapes from a single image, which cannot reproduce shape details in occluded body regions. The missing information of the monocular input can be recovered by using multiple views captured from multiple cameras. However, multi-view reconstruction methods necessitate accurately calibrated and registered images, which can be challenging to obtain in real-world scenarios. Given a target RGB image and a collection of multiple uncalibrated and unregistered images of the same individual, acquired using a single camera, we propose a novel framework to generate complete 3D human shapes. We introduce a novel module to generate 2D multi-view normal maps of the person registered with the target input image. The module consists of body part-based reference selection and body part-based registration. The generated 2D normal maps are then processed by a multi-view attention-based neural implicit model that estimates an implicit representation of the 3D shape, ensuring the reproduction of details in both observed and occluded regions. Extensive experiments demonstrate that the proposed approach estimates higher quality details in the non-visible regions of the 3D clothed human shapes compared to related methods, without using parametric models.

Creating 3D semantic reconstructions of environments is fundamental to many applications, especially when related to autonomous agent operation (e.g., goal-oriented navigation or object interaction and manipulation). Commonly, 3D semantic reconstruction systems capture the entire scene in the same level of detail. However, certain tasks (e.g., object interaction) require a fine-grained and high-resolution map, particularly if the objects to interact are of small size or intricate geometry. In recent practice, this leads to the entire map being in the same high-quality resolution, which results in increased computational and storage costs. To address this challenge, we propose MAP-ADAPT, a real-time method for quality-adaptive semantic 3D reconstruction using RGBD frames. MAP-ADAPT is the first adaptive semantic 3D mapping algorithm that, unlike prior work, generates directly a single map with regions of different quality based on both the semantic information and the geometric complexity of the scene. Leveraging a semantic SLAM pipeline for pose and semantic estimation, we achieve comparable or superior results to state-of-the-art methods on synthetic and real-world data, while significantly reducing storage and computation requirements. Code is available at https://​map-adapt.​github.​io/​.

The goal of this paper is to discover, segment, and track independently moving objects in complex visual scenes. Previous approaches have explored the use of optical flow for motion segmentation, leading to imperfect predictions due to partial motion, background distraction, and object articulations and interactions. To address this issue, we introduce an appearance-based refinement method that leverages temporal consistency in video streams to correct inaccurate flow-based proposals. Our approach involves a sequence-level selection mechanism that identifies accurate flow-predicted masks as exemplars, and an object-centric architecture that refines problematic masks based on exemplar information. The model is pre-trained on synthetic data and then adapted to real-world videos in a self-supervised manner, eliminating the need for human annotations. Its performance is evaluated on multiple video segmentation benchmarks, including DAVIS, YouTubeVOS, SegTrackv2, and FBMS-59. We achieve competitive performance on single-object segmentation, while significantly outperforming existing models on the more challenging problem of multi-object segmentation. Finally, we investigate the benefits of using our model as a prompt for the per-frame Segment Anything Model.

In semi-supervised semantic segmentation, a model is trained with a limited number of labeled images along with a large corpus of unlabeled images to reduce the high annotation effort. While previous methods are able to learn good segmentation boundaries, they are prone to confuse classes with similar visual appearance due to the limited supervision. On the other hand, vision-language models (VLMs) are able to learn diverse semantic knowledge from image-caption datasets but produce noisy segmentation due to the image-level training. In SemiVL, we newly propose to integrate rich priors from VLM pre-training into semi-supervised semantic segmentation to learn better semantic decision boundaries. To adapt the VLM from global to local reasoning, we introduce a spatial fine-tuning strategy for label-efficient learning. Further, we design a language-guided decoder to jointly reason over vision and language. Finally, we propose to handle inherent ambiguities in class labels by instructing the model with language guidance in the form of class definitions. We evaluate SemiVL on 4 semantic segmentation datasets, where it significantly outperforms previous semi-supervised methods. For instance, SemiVL improves the state of the art by +13.5 mIoU on COCO with 232 annotated images and by +6.1 mIoU on Pascal VOC with 92 annotated images. Project page: github.​com/​google-research/​semivl

Pre-trained vision-language models (VLMs) have enabled significant progress in open vocabulary computer vision tasks such as image classification, object detection and image segmentation. Some recent works have focused on extending VLMs to open vocabulary single label action classification in videos. However, previous methods fall short in holistic video understanding which requires the ability to simultaneously recognize multiple actions and entities e.g., objects in the video in an open vocabulary setting. We formulate this problem as open vocabulary multi-label video classification and propose a method to adapt a pre-trained VLM such as CLIP to solve this task. We leverage large language models (LLMs) to provide semantic guidance to the VLM about class labels to improve its open vocabulary performance with two key contributions. First, we propose an end-to-end trainable architecture that learns to prompt an LLM to generate soft attributes for the CLIP text-encoder to enable it to recognize novel classes. Second, we integrate a temporal modeling module into CLIP’s vision encoder to effectively model the spatio-temporal dynamics of video concepts as well as propose a novel regularized finetuning technique to ensure strong open vocabulary classification performance in the video domain. Our extensive experimentation showcases the efficacy of our approach on multiple benchmark datasets.

Noisy few-shot learning (NFSL) presents novel challenges primarily due to the interplay between noisy labels and limited training data. While data cleansing offers a viable solution to address noisy labels in the general learning settings, it exacerbates information loss in FSL due to limited training data, resulting in inadequate model training. To best recover the underlying task manifold corrupted by the noisy labels, we resort to learning from uniquely designed unsupervised auxiliary tasks to compensate for information loss. Using unsupervised tasks can effectively avoid additional annotation costs and minimize the risk of introducing additional label noises. However, a randomly constructed unsupervised task may misguide the model to learn sample-specific features that are likely to compromise the primary few-shot learning task due to the noisy weak learning signals. We propose to conduct novel auxiliary task selection to ensure the intra-diversity among the unlabeled samples within a task. Domain invariant features are then learned from carefully constructed auxiliary tasks to best recover the original data manifold. We conduct a theoretical analysis to derive novel generalization bounds for learning with auxiliary tasks. Extensive experiments are conducted to demonstrate that our method outperforms existing noisy few-shot learning methods under various in-domain and cross-domain few-shot classification benchmarks.

This paper presents a novel approach for reconstructing dynamic radiance fields from monocular videos. We integrate kinematics with dynamic radiance fields, bridging the gap between the sparse nature of monocular videos and the real-world physics. Our method introduces the kinematic field, capturing motion through kinematic quantities: velocity, acceleration, and jerk. The kinematic field is jointly learned with the dynamic radiance field by minimizing the photometric loss without motion ground truth. We further augment our method with physics-driven regularizers grounded in kinematics. We propose physics-driven regularizers that ensure the physical validity of predicted kinematic quantities, including advective acceleration and jerk. Additionally, we control the motion trajectory based on rigidity equations formed with the predicted kinematic quantities. In experiments, our method outperforms the state-of-the-arts by capturing physical motion patterns within challenging real-world monocular videos.

Unsupervised Domain Adaptation (UDA) is the task of bridging the domain gap between a labeled source domain, e.g., synthetic data, and an unlabeled target domain. We observe that current UDA methods show inferior results on fine structures and tend to oversegment objects with ambiguous appearance. To address these shortcomings, we propose to leverage geometric information, i.e., depth predictions, as depth discontinuities often coincide with segmentation boundaries. We show that naively incorporating depth into current UDA methods does not fully exploit the potential of this complementary information. To this end, we present MICDrop, which learns a joint feature representation by masking image encoder features while inversely masking depth encoder features. With this simple yet effective complementary masking strategy, we enforce the use of both modalities when learning the joint feature representation. To aid this process, we propose a feature fusion module to improve both global as well as local information sharing while being robust to errors in the depth predictions. We show that our method can be plugged into various recent UDA methods and consistently improve results across standard UDA benchmarks, obtaining new state-of-the-art performances. Project Page: https://​github.​com/​ly-muc/​MICDrop.

Procedural videos, exemplified by recipe demonstrations, are instrumental in conveying step-by-step instructions. However, understanding such videos is challenging as it involves the precise localization of steps and the generation of textual instructions. Manually annotating steps and writing instructions is costly, which limits the size of current datasets and hinders effective learning. Leveraging large but noisy video-transcript datasets for pre-training can boost performance but demands significant computational resources. Furthermore, transcripts contain irrelevant content and differ in style from human-written instructions. To mitigate these issues, we propose a novel technique, Sieve  & Swap, to automatically generate high-quality training data for the recipe domain: (i) Sieve: filters irrelevant transcripts and (ii) Swap: acquires high-quality text by replacing transcripts with human-written instruction from a text-only recipe dataset. The resulting dataset is three orders of magnitude smaller than current web-scale datasets but enables efficient training of large-scale models. Alongside Sieve  & Swap, we propose Procedure Transformer (ProcX), a model for end-to-end step localization and instruction generation for procedural videos. When pre-trained on our curated dataset, this model achieves state-of-the-art performance on YouCook2 and Tasty while using a fraction of the training data. We have released code and dataset. (https://​github.​com/​anilbatra2185/​sns_​procx).

Knowledge distillation (KD) has emerged as a promising technique in deep learning, typically employed to enhance a compact student network through learning from their high-performance but more complex teacher variant. When applied in the context of image super-resolution, most KD approaches are modified versions of methods developed for other computer vision tasks, which are based on training strategies with a single teacher and simple loss functions. In this paper, we propose a novel Multi-Teacher Knowledge Distillation (MTKD) framework specifically for image super-resolution. It exploits the advantages of multiple teachers by combining and enhancing the outputs of these teacher models, which then guides the learning process of the compact student network. To achieve more effective learning performance, we have also developed a new wavelet-based loss function for MTKD, which can better optimize the training process by observing differences in both the spatial and frequency domains. We fully evaluate the effectiveness of the proposed method by comparing it to five commonly used KD methods for image super-resolution based on three popular network architectures. The results show that the proposed MTKD method achieves evident improvements in super-resolution performance, up to 0.46 dB (based on PSNR), over state-of-the-art KD approaches across different network structures. The source code of MTKD will be made available https://​github.​com/​YuxuanJJ/​MTKD for public evaluation.

Large pre-trained Vision-Language Models (VLMs) have become ubiquitous foundational components of other models and downstream tasks. Although powerful, our empirical results reveal that such models might not be able to identify fine-grained concepts. Specifically, the explanations of VLMs with respect to fine-grained concepts are entangled and mislocalized. To address this issue, we propose to DisEntAngle and Localize (DEAL) the concept-level explanations for VLMs without human annotations. The key idea is encouraging the concept-level explanations to be distinct while maintaining consistency with category-level explanations. We conduct extensive experiments and ablation studies on a wide range of benchmark datasets and vision-language models. Our empirical results demonstrate that the proposed method significantly improves the concept-level explanations of the model in terms of disentanglability and localizability. Surprisingly, the improved explainability alleviates the model’s reliance on spurious correlations, which further benefits the prediction accuracy.

Despite the abundant availability and content richness for video data, its high-dimensionality poses challenges for video research. Recent advancements have explored the implicit representation for videos using neural networks, demonstrating strong performance in applications such as video compression and enhancement. However, the prolonged encoding time remains a persistent challenge for video Implicit Neural Representations (INRs). In this paper, we focus on improving the speed of video encoding and decoding within implicit representations. We introduce two key components: NeRV-Enc, a transformer-based hyper-network for fast encoding; and NeRV-Dec, a parallel decoder for efficient video loading. NeRV-Enc achieves an impressive speed-up of \(\mathbf {10^4\times }\) by eliminating gradient-based optimization. Meanwhile, NeRV-Dec simplifies video decoding, outperforming conventional codecs with a loading speed \(\mathbf {11\times }\) faster, and surpassing RAM loading with pre-decoded videos (\(\mathbf {2.5\times }\) faster while being \(\mathbf {65\times }\) smaller in size).

We present Surf-D, a novel method for generating high-quality 3D shapes as Surfaces with arbitrary topologies using Diffusion models. Previous methods explored shape generation with different representations and they suffer from limited topologies and poor geometry details. To generate high-quality surfaces of arbitrary topologies, we use the Unsigned Distance Field (UDF) as our surface representation to accommodate arbitrary topologies. Furthermore, we propose a new pipeline that employs a point-based AutoEncoder to learn a compact and continuous latent space for accurately encoding UDF and support high-resolution mesh extraction. We further show that our new pipeline significantly outperforms the prior approaches to learning the distance fields, such as the grid-based AutoEncoder, which is not scalable and incapable of learning accurate UDF. In addition, we adopt a curriculum learning strategy to efficiently embed various surfaces. With the pretrained shape latent space, we employ a latent diffusion model to acquire the distribution of various shapes. Extensive experiments are presented on using Surf-D for unconditional generation, category conditional generation, image conditional generation, and text-to-shape tasks. The experiments demonstrate the superior performance of Surf-D in shape generation across multiple modalities as conditions. Visit our project page at https://​yzmblog.​github.​io/​projects/​SurfD/​.

Training gaze following models requires a large number of images with gaze target coordinates annotated by human annotators, which is a laborious and inherently ambiguous process. We propose the first semi-supervised method for gaze following by introducing two novel priors to the task. We obtain the first prior using a large pretrained Visual Question Answering (VQA) model, where we compute Grad-CAM heatmaps by ‘prompting’ the VQA model with a gaze following question. These heatmaps can be noisy and not suited for use in training. The need to refine these noisy annotations leads us to incorporate a second prior. We utilize a diffusion model trained on limited human annotations and modify the reverse sampling process to refine the Grad-CAM heatmaps. By tuning the diffusion process we achieve a trade-off between the human annotation prior and the VQA heatmap prior, which retains the useful VQA prior information while exhibiting similar properties to the training data distribution. Our method outperforms simple pseudo-annotation generation baselines on the GazeFollow image dataset. More importantly, our pseudo-annotation strategy, applied to a widely used supervised gaze following model (VAT), reduces the annotation need by 50%. Our method also performs the best on the VideoAttentionTarget dataset. Code is available at https://​github.​com/​cvlab-stonybrook/​GCDR-Gaze.​git.

Advancements in open-sourced text-to-image models and fine-tuning methods have led to the increasing risk of malicious adaptation, i.e., fine-tuning to generate harmful/unauthorized content. Recent works, e.g., Glaze or MIST, have developed data-poisoning techniques which protect the data against adaptation methods. In this work, we consider an alternative paradigm for protection. We propose to “immunize” the model by learning model parameters that are difficult for the adaptation methods when fine-tuning malicious content; in short IMMA. Specifically, IMMA should be applied before the release of the model weights to mitigate these risks. Empirical results show IMMA’s effectiveness against malicious adaptations, including mimicking the artistic style and learning of inappropriate/unauthorized content, over three adaptation methods: LoRA, Textual-Inversion, and DreamBooth. The code is available at https://​github.​com/​amberyzheng/​IMMA.

This paper introduces Motion-oriented Compositional Neu-ral Radiance Fields (MoCo-NeRF), a framework designed to perform free-viewpoint rendering of monocular human videos via novel non-rigid motion modeling approach. In the context of dynamic clothed humans, complex cloth dynamics generate non-rigid motions that are intrinsically distinct from skeletal articulations and critically important for the rendering quality. The conventional approach models non-rigid motions as spatial (3D) deviations in addition to skeletal transformations. However, it is either time-consuming or challenging to achieve optimal quality due to its high learning complexity without a direct supervision. To target this problem, we propose a novel approach of modeling non-rigid motions as radiance residual fields to benefit from more direct color supervision in the rendering and utilize the rigid radiance fields as a prior to reduce the complexity of the learning process. Our approach utilizes a single multiresolution hash encoding (MHE) to concurrently learn the canonical T-pose representation from rigid skeletal motions and the radiance residual field for non-rigid motions. Additionally, to further improve both training efficiency and usability, we extend MoCo-NeRF to support simultaneous training of multiple subjects within a single framework, thanks to our effective design for modeling non-rigid motions. This scalability is achieved through the integration of a global MHE and learnable identity codes in addition to multiple local MHEs. We present extensive results on ZJU-MoCap and MonoCap, clearly demonstrating state-of-the-art performance in both single- and multi-subject settings. The code and model will be made publicly available at the project page: https://​stevejaehyeok.​github.​io/​publications/​moco-nerf.