CVPR 2024 Papers

Self-attention mechanism is the key of the Transformer but often criticized for its computation demands. Previous token pruning works motivate their methods from the view of computation redundancy but still need to load the full network and require same memory costs. This paper introduces a novel strategy that simplifies vision transformers and reduces computational load through the selective removal of non-essential attention layers guided by entropy considerations. We identify that regarding the attention layer in bottom blocks their subsequent MLP layers i.e. two feed-forward layers can elicit the same entropy quantity. Meanwhile the accompanied MLPs are under-exploited since they exhibit smaller feature entropy compared to those MLPs in the top blocks. Therefore we propose to integrate the uninformative attention layers into their subsequent counterparts by degenerating them into identical mapping yielding only MLP in certain transformer blocks. Experimental results on ImageNet-1k show that the proposed method can remove 40% attention layer of DeiT-B improving throughput and memory bound without performance compromise.

Vision-language generative AI has demonstrated remarkable promise for empowering cross-modal scene understanding of autonomous driving and high-definition (HD) map systems. However current benchmark datasets lack multi-modal point cloud image and language data pairs. Recent approaches utilize visual instruction learning and cross-modal prompt engineering to expand vision-language models into this domain. In this paper we propose a new vision-language benchmark that can be used to finetune traffic and HD map domain-specific foundation models. Specifically we annotate and leverage large-scale broad-coverage traffic and map data extracted from huge HD map annotations and use CLIP and LLaMA-2 / Vicuna to finetune a baseline model with instruction-following data. Our experimental results across various algorithms reveal that while visual instruction-tuning large language models (LLMs) can effectively learn meaningful representations from MAPLM-QA there remains significant room for further advancements. To facilitate applying LLMs and multi-modal data into self-driving research we will release our visual-language QA data and the baseline models at GitHub.com/LLVM-AD/MAPLM.

Understanding how we grasp objects with our hands has important applications in areas like robotics and mixed reality. However this challenging problem requires accurate modeling of the contact between hands and objects.To capture grasps existing methods use skeletons meshes or parametric models that does not represent hand shape accurately resulting in inaccurate contacts. We present MANUS a method for Markerless Hand-Object Grasp Capture using Articulated 3D Gaussians. We build a novel articulated 3D Gaussians representation that extends 3D Gaussian splatting for high-fidelity representation of articulating hands. Since our representation uses Gaussian primitives optimized from the multi-view pixel-aligned losses it enables us to efficiently and accurately estimate contacts between the hand and the object. For the most accurate results our method requires tens of camera views that current datasets do not provide. We therefore build MANUS-Grasps a new dataset that contains hand-object grasps viewed from 50+ cameras across 30+ scenes 3 subjects and comprising over 7M frames. In addition to extensive qualitative results we also show that our method outperforms others on a quantitative contact evaluation method that uses paint transfer from the object to the hand.

The recent advancements in text-to-3D generation mark a significant milestone in generative models unlocking new possibilities for creating imaginative 3D assets across various real-world scenarios. While recent advancements in text-to-3D generation have shown promise they often fall short in rendering detailed and high-quality 3D models. This problem is especially prevalent as many methods base themselves on Score Distillation Sampling (SDS). This paper identifies a notable deficiency in SDS that it brings inconsistent and low-quality updating direction for the 3D model causing the over-smoothing effect. To address this we propose a novel approach called Interval Score Matching (ISM). ISM employs deterministic diffusing trajectories and utilizes interval-based score matching to counteract over-smoothing. Furthermore we incorporate 3D Gaussian Splatting into our text-to-3D generation pipeline. Extensive experiments show that our model largely outperforms the state-of-the-art in quality and training efficiency.

Rotation invariance is an important requirement for point shape analysis. To achieve this current state-of-the-art methods attempt to construct the local rotation-invariant representation through learning or defining the local reference frame (LRF). Although efficient these LRF-based methods suffer from perturbation of local geometric relations resulting in suboptimal local rotation invariance. To alleviate this issue we propose a Local-consistent Transformation (LocoTrans) learning strategy. Specifically we first construct the local-consistent reference frame (LCRF) by considering the symmetry of the two axes in LRF. In comparison with previous LRFs our LCRF is able to preserve local geometric relationships better through performing local-consistent transformation. However as the consistency only exists in local regions the relative pose information is still lost in the intermediate layers of the network. We mitigate such a relative pose issue by developing a relative pose recovery (RPR) module. RPR aims to restore the relative pose between adjacent transformed patches. Equipped with LCRF and RPR our LocoTrans is capable of learning local-consistent transformation and preserving local geometry which benefits rotation invariance learning. Competitive performance under arbitrary rotations on both shape classification and part segmentation tasks and ablations can demonstrate the effectiveness of our method. Code will be available publicly at https://github.com/wdttt/LocoTrans.

Research into dynamic 3D scene understanding has primarily focused on short-term change tracking from dense observations while little attention has been paid to long-term changes with sparse observations. We address this gap with MoRE a novel approach for multi-object relocalization and reconstruction in evolving environments. We view these environments as Living Scenes and consider the problem of transforming scans taken at different points in time into a 3D reconstruction of the object instances whose accuracy and completeness increase over time. At the core of our method lies an SE(3) equivariant representation in a single encoder-decoder network trained on synthetic data. This representation enables us to seamlessly tackle instance matching registration and reconstruction. We also introduce a joint optimization algorithm that facilitates the accumulation of point clouds originating from the same instance across multiple scans taken at different points in time. We validate our method on synthetic and real-world data and demonstrate state-of-the-art performance in both end-to-end performance and individual subtasks.

Photorealistic simulation plays a crucial role in applications such as autonomous driving where advances in neural radiance fields (NeRFs) may allow better scalability through the automatic creation of digital 3D assets. However reconstruction quality suffers on street scenes due to largely collinear camera motions and sparser samplings at higher speeds. On the other hand the application often demands rendering from camera views that deviate from the inputs to accurately simulate behaviors like lane changes. In this paper we propose several insights that allow a better utilization of Lidar data to improve NeRF quality on street scenes. First our framework learns a geometric scene representation from Lidar which are fused with the implicit grid-based representation for radiance decoding thereby supplying stronger geometric information offered by explicit point cloud. Second we put forth a robust occlusion-aware depth supervision scheme which allows utilizing densified Lidar points by accumulation. Third we generate augmented training views from Lidar points for further improvement. Our insights translate to largely improved novel view synthesis under real driving scenes.

LiDAR localization is a fundamental task in robotics and computer vision which estimates the pose of a LiDAR point cloud within a global map. Scene Coordinate Regression (SCR) has demonstrated state-of-the-art performance in this task. In SCR a scene is represented as a neural network which outputs the world coordinates for each point in the input point cloud. However SCR treats all points equally during localization ignoring the fact that not all objects are beneficial for localization. For example dynamic objects and repeating structures often negatively impact SCR. To address this problem we introduce LiSA the first method that incorporates semantic awareness into SCR to boost the localization robustness and accuracy. To avoid extra computation or network parameters during inference we distill the knowledge from a segmentation model to the original SCR network. Experiments show the superior performance of LiSA on standard LiDAR localization benchmarks compared to state-of-the-art methods. Applying knowledge distillation not only preserves high efficiency but also achieves higher localization accuracy than introducing extra semantic segmentation modules. We also analyze the benefit of semantic information for LiDAR localization. Our code is released at https://github.com/Ybchun/LiSA.

Although neural radiance fields (NeRFs) have achieved triumphs in image novel view synthesis (NVS) LiDAR NVS remains largely unexplored. Previous LiDAR NVS methods employ a simple shift from image NVS methods while ignoring the dynamic nature and the large-scale reconstruction problem of LiDAR point clouds. In light of this we propose LiDAR4D a differentiable LiDAR-only framework for novel space-time LiDAR view synthesis. In consideration of the sparsity and large-scale characteristics we design a 4D hybrid representation combined with multi-planar and grid features to achieve effective reconstruction in a coarse-to-fine manner. Furthermore we introduce geometric constraints derived from point clouds to improve temporal consistency. For the realistic synthesis of LiDAR point clouds we incorporate the global optimization of ray-drop probability to preserve cross-region patterns. Extensive experiments on KITTI-360 and NuScenes datasets demonstrate the superiority of our method in accomplishing geometry-aware and time-consistent dynamic reconstruction. Codes are available at https://github.com/ispc-lab/LiDAR4D.

In this paper we present LiDAR-Net a new real-scanned indoor point cloud dataset containing nearly 3.6 billion precisely point-level annotated points covering an expansive area of 30000m^2. It encompasses three prevalent daily environments including learning scenes working scenes and living scenes. LiDAR-Net is characterized by its non-uniform point distribution e.g. scanning holes and scanning lines. Additionally it meticulously records and annotates scanning anomalies including reflection noise and ghost. These anomalies stem from specular reflections on glass or metal as well as distortions due to moving persons. LiDAR-Net's realistic representation of non-uniform distribution and anomalies significantly enhances the training of deep learning models leading to improved generalization in practical applications. We thoroughly evaluate the performance of state-of-the-art algorithms on LiDAR-Net and provide a detailed analysis of the results. Crucially our research identifies several fundamental challenges in understanding indoor point clouds contributing essential insights to future explorations in this field. Our dataset can be found online: http://lidar-net.njumeta.com

We propose a voxel-based optimization framework ReVoRF for few-shot radiance fields that strategically addresses the unreliability in pseudo novel view synthesis. Our method pivots on the insight that relative depth relationships within neighboring regions are more reliable than the absolute color values in disoccluded areas. Consequently we devise a bilateral geometric consistency loss that carefully navigates the trade-off between color fidelity and geometric accuracy in the context of depth consistency for uncertain regions. Moreover we present a reliability-guided learning strategy to discern and utilize the variable quality across synthesized views complemented by a reliability-aware voxel smoothing algorithm that smoothens the transition between reliable and unreliable data patches. Our approach allows for a more nuanced use of all available data promoting enhanced learning from regions previously considered unsuitable for high-quality reconstruction. Extensive experiments across diverse datasets reveal that our approach attains significant gains in efficiency and accuracy delivering rendering speeds of 3 FPS 7 mins to train a 360deg scene and a 5% improvement in PSNR over existing few-shot methods. Code is available at https://github.com/HKCLynn/ReVoRF

We introduce SynCLR a novel approach for learning visual representations exclusively from synthetic images without any real data. We synthesize a large dataset of image captions using LLMs then use an off-the-shelf text-to-image model to generate multiple images corresponding to each synthetic caption. We perform visual representation learning on these synthetic images via contrastive learning treating images sharing the same caption as positive pairs. The resulting representations demonstrate remarkable transferability competing favorably with other general-purpose visual representation learners such as CLIP and DINO v2 in image classification tasks. Furthermore in dense prediction tasks such as semantic segmentation SynCLR outperforms previous self-supervised methods by a significant margin e.g. improving over MAE and iBOT by 5.0 and 3.1 mIoU on ADE20k for ViT-B/16.

Establishing accurate 3D correspondences between shapes stands as a pivotal challenge with profound implications for computer vision and robotics. However existing self-supervised methods for this problem assume perfect input shape alignment restricting their real-world applicability. In this work we introduce a novel self-supervised Rotation-Invariant 3D correspondence learner with Local Shape Transform dubbed RIST that learns to establish dense correspondences between shapes even under challenging intra-class variations and arbitrary orientations. Specifically RIST learns to dynamically formulate an SO(3)-invariant local shape transform for each point which maps the SO(3)-equivariant global shape descriptor of the input shape to a local shape descriptor. These local shape descriptors are provided as inputs to our decoder to facilitate point cloud self- and cross-reconstruction. Our proposed self-supervised training pipeline encourages semantically corresponding points from different shapes to be mapped to similar local shape descriptors enabling RIST to establish dense point-wise correspondences. RIST demonstrates state-of-the-art performances on 3D part label transfer and semantic keypoint transfer given arbitrarily rotated point cloud pairs outperforming existing methods by significant margins.

Multi-instance point cloud registration estimates the poses of multiple instances of a model point cloud in a scene point cloud. Extracting accurate point correspondences is to the center of the problem. Existing approaches usually treat the scene point cloud as a whole overlooking the separation of instances. Therefore point features could be easily polluted by other points from the background or different instances leading to inaccurate correspondences oblivious to separate instances especially in cluttered scenes. In this work we propose MIRETR Multi-Instance REgistration TRansformer a coarse-to-fine approach to the extraction of instance-aware correspondences. At the coarse level it jointly learns instance-aware superpoint features and predicts per-instance masks. With instance masks the influence from outside of the instance being concerned is minimized such that highly reliable superpoint correspondences can be extracted. The superpoint correspondences are then extended to instance candidates at the fine level according to the instance masks. At last an efficient candidate selection and refinement algorithm is devised to obtain the final registrations. Extensive experiments on three public benchmarks demonstrate the efficacy of our approach. In particular MIRETR outperforms the state of the arts by 16.6 points on F1 score on the challenging ROBI benchmark. Code and models are available at https://github.com/zhiyuanYU134/MIRETR

Recent works in implicit representations such as Neural Radiance Fields (NeRF) have advanced the generation of realistic and animatable head avatars from video sequences. These implicit methods are still confronted by visual artifacts and jitters since the lack of explicit geometric constraints poses a fundamental challenge in accurately modeling complex facial deformations. In this paper we introduce Dynamic Tetrahedra (DynTet) a novel hybrid representation that encodes explicit dynamic meshes by neural networks to ensure geometric consistency across various motions and viewpoints. DynTet is parameterized by the coordinate-based networks which learn signed distance deformation and material texture anchoring the training data into a predefined tetrahedra grid. Leveraging Marching Tetrahedra DynTet efficiently decodes textured meshes with a consistent topology enabling fast rendering through a differentiable rasterizer and supervision via a pixel loss. To enhance training efficiency we incorporate classical 3D Morphable Models to facilitate geometry learning and define a canonical space for simplifying texture learning. These advantages are readily achievable owing to the effective geometric representation employed in DynTet. Compared with prior works DynTet demonstrates significant improvements in fidelity lip synchronization and real-time performance according to various metrics. Beyond producing stable and visually appealing synthesis videos our method also outputs the dynamic meshes which is promising to enable many emerging applications. Code is available at https://github.com/zhangzc21/DynTet.

We introduce a new attention mechanism dubbed structural self-attention (StructSA) that leverages rich correlation patterns naturally emerging in key-query interactions of attention. StructSA generates attention maps by recognizing space-time structures of key-query correlations via convolution and uses them to dynamically aggregate local contexts of value features. This effectively leverages rich structural patterns in images and videos such as scene layouts object motion and inter-object relations.Using StructSA as a main building block we develop the structural vision transformer (StructViT) and evaluate its effectiveness on both image and video classification tasks achieving state-of-the-art results on ImageNet-1K Kinetics-400 Something-Something V1 & V2 Diving-48 and FineGym.

We propose an unsupervised method for parsing large 3D scans of real-world scenes with easily-interpretable shapes. This work aims to provide a practical tool for analyzing 3D scenes in the context of aerial surveying and mapping without the need for user annotations. Our approach is based on a probabilistic reconstruction model that decomposes an input 3D point cloud into a small set of learned prototypical 3D shapes. The resulting reconstruction is visually interpretable and can be used to perform unsupervised instance and low-shot semantic segmentation of complex scenes. We demonstrate the usefulness of our model on a novel dataset of seven large aerial LiDAR scans from diverse real-world scenarios. Our approach outperforms state-of-the-art unsupervised methods in terms of decomposition accuracy while remaining visually interpretable. Our code and dataset are available at https://romainloiseau.fr/learnable-earth-parser/.

Neural radiance field (NeRF) is an emerging technique for 3D scene reconstruction and modeling. However current NeRF-based methods are limited in the capabilities of adding or removing objects. This paper fills the aforementioned gap by proposing a new language-driven method for object manipulation in NeRFs through dataset updates. Specifically to insert an object represented by a set of multi-view images into a background NeRF we use a text-to-image diffusion model to blend the object into the given background across views. The generated images are then used to update the NeRF so that we can render view-consistent images of the object within the background. To ensure view consistency we propose a dataset update strategy that prioritizes the radiance field training based on camera poses in a pose-ordered manner. We validate our method in two case studies: object insertion and object removal. Experimental results show that our method can generate photo-realistic results and achieves state-of-the-art performance in NeRF editing.

Open-vocabulary querying in 3D space is challenging but essential for scene understanding tasks such as object localization and segmentation. Language-embedded scene representations have made progress by incorporating language features into 3D spaces. However their efficacy heavily depends on neural networks that are resource-intensive in training and rendering. Although recent 3D Gaussians offer efficient and high-quality novel view synthesis directly embedding language features in them leads to prohibitive memory usage and decreased performance. In this work we introduce Language Embedded 3D Gaussians a novel scene representation for open-vocabulary query tasks. Instead of embedding high-dimensional raw semantic features on 3D Gaussians we propose a dedicated quantization scheme that drastically alleviates the memory requirement and a novel embedding procedure that achieves smoother yet high accuracy query countering the multi-view feature inconsistencies and the high-frequency inductive bias in point-based representations. Our comprehensive experiments show that our representation achieves the best visual quality and language querying accuracy across current language-embedded representations while maintaining real-time rendering frame rates on a single desktop GPU.