Search

Reconstructing scenes and tracking motion are two sides of the same coin. Tracking points allow for geometric reconstruction [14], while geometric reconstruction of (dynamic) scenes allows for 3D tracking of points over time [24, 39]. The latter was recently also exploited for 2D point tracking to overcome occlusion ambiguities by lifting tracking directly into 3D [38]. However, above approaches either require offline processing or multi-view camera setups both unrealistic for real-world applications like robot navigation or mixed reality. We target the challenge of online 2D and 3D point tracking from unposed monocular camera input introducing Dynamic Online Monocular Reconstruction (DynOMo). We leverage 3D Gaussian splatting to reconstruct dynamic scenes in an online fashion. Our approach extends 3D Gaussians to capture new content and object motions while estimating camera movements from a single RGB frame. DynOMo stands out by enabling emergence of point trajectories through robust image feature reconstruction and a novel similarity-enhanced regularization term, without requiring any correspondence-level supervision. It sets the first baseline for online point tracking with monocular unposed cameras, achieving performance on par with existing methods. We aim to inspire the community to advance online point tracking and reconstruction, expanding the applicability to diverse real-world scenarios.

Autonomous motorsports aim to replicate the human racecar driver with software and sensors. As in traditional motorsports, Autonomous Racing Vehicles (ARVs) are pushed to their handling limits in multi-agent scenarios at extremely high ($\geq 150mph$) speeds. This Operational Design Domain (ODD) presents unique challenges across the autonomy stack. The Indy Autonomous Challenge (IAC) is an international competition aiming to advance autonomous vehicle development through ARV competitions. While far from challenging what a human racecar driver can do, the IAC is pushing the state of the art by facilitating full-sized ARV competitions. This paper details the MIT-Pitt-RW Team's approach to autonomous racing in the IAC. In this work, we present our modular and fast approach to agent detection, motion planning and controls to create an autonomy stack. We also provide analysis of the performance of the software stack in single and multi-agent scenarios for rapid deployment in a fast-paced competition environment. We also cover what did and did not work when deployed on a physical system the Dallara AV-21 platform and potential improvements to address these shortcomings. Finally, we convey lessons learned and discuss limitations and future directions for improvement., Comment: Published in Journal of Field Robotics

In this paper, we present a method for dynamic surface reconstruction of large-scale urban scenes from LiDAR. Depth-based reconstructions tend to focus on small-scale objects or large-scale SLAM reconstructions that treat moving objects as outliers. We take a holistic perspective and optimize a compositional model of a dynamic scene that decomposes the world into rigidly moving objects and the background. To achieve this, we take inspiration from recent novel view synthesis methods and pose the reconstruction problem as a global optimization, minimizing the distance between our predicted surface and the input LiDAR scans. We show how this global optimization can be decomposed into registration and surface reconstruction steps, which are handled well by off-the-shelf methods without any re-training. By careful modeling of continuous-time motion, our reconstructions can compensate for the rolling shutter effects of rotating LiDAR sensors. This allows for the first system (to our knowledge) that properly motion compensates LiDAR scans for rigidly-moving objects, complementing widely-used techniques for motion compensation of static scenes. Beyond pursuing dynamic reconstruction as a goal in and of itself, we also show that such a system can be used to auto-label partially annotated sequences and produce ground truth annotation for hard-to-label problems such as depth completion and scene flow.

While text-to-visual models now produce photo-realistic images and videos, they struggle with compositional text prompts involving attributes, relationships, and higher-order reasoning such as logic and comparison. In this work, we conduct an extensive human study on GenAI-Bench to evaluate the performance of leading image and video generation models in various aspects of compositional text-to-visual generation. We also compare automated evaluation metrics against our collected human ratings and find that VQAScore -- a metric measuring the likelihood that a VQA model views an image as accurately depicting the prompt -- significantly outperforms previous metrics such as CLIPScore. In addition, VQAScore can improve generation in a black-box manner (without finetuning) via simply ranking a few (3 to 9) candidate images. Ranking by VQAScore is 2x to 3x more effective than other scoring methods like PickScore, HPSv2, and ImageReward at improving human alignment ratings for DALL-E 3 and Stable Diffusion, especially on compositional prompts that require advanced visio-linguistic reasoning. We will release a new GenAI-Rank benchmark with over 40,000 human ratings to evaluate scoring metrics on ranking images generated from the same prompt. Lastly, we discuss promising areas for improvement in VQAScore, such as addressing fine-grained visual details. We will release all human ratings (over 80,000) to facilitate scientific benchmarking of both generative models and automated metrics., Comment: We open-source our dataset, model, and code at: https://linzhiqiu.github.io/papers/genai_bench ; Project page: https://linzhiqiu.github.io/papers/genai_bench ; GenAI-Bench was first introduced in arxiv:2404.01291. This article extends it with an additional GenAI-Rank benchmark.

Motion planning is crucial for safe navigation in complex urban environments. Historically, motion planners (MPs) have been evaluated with procedurally-generated simulators like CARLA. However, such synthetic benchmarks do not capture real-world multi-agent interactions. nuPlan, a recently released MP benchmark, addresses this limitation by augmenting real-world driving logs with closed-loop simulation logic, effectively turning the fixed dataset into a reactive simulator. We analyze the characteristics of nuPlan's recorded logs and find that each city has its own unique driving behaviors, suggesting that robust planners must adapt to different environments. We learn to model such unique behaviors with BehaviorNet, a graph convolutional neural network (GCNN) that predicts reactive agent behaviors using features derived from recently-observed agent histories; intuitively, some aggressive agents may tailgate lead vehicles, while others may not. To model such phenomena, BehaviorNet predicts the parameters of an agent's motion controller rather than directly predicting its spacetime trajectory (as most forecasters do). Finally, we present AdaptiveDriver, a model-predictive control (MPC) based planner that unrolls different world models conditioned on BehaviorNet's predictions. Our extensive experiments demonstrate that AdaptiveDriver achieves state-of-the-art results on the nuPlan closed-loop planning benchmark, improving over prior work by 2% on Test-14 Hard R-CLS, and generalizes even when evaluated on never-before-seen cities., Comment: Project Page: https://arunbalajeev.github.io/world_models_planning/world_model_paper.html

State-of-the-art 3D object detectors are often trained on massive labeled datasets. However, annotating 3D bounding boxes remains prohibitively expensive and time-consuming, particularly for LiDAR. Instead, recent works demonstrate that self-supervised pre-training with unlabeled data can improve detection accuracy with limited labels. Contemporary methods adapt best-practices for self-supervised learning from the image domain to point clouds (such as contrastive learning). However, publicly available 3D datasets are considerably smaller and less diverse than those used for image-based self-supervised learning, limiting their effectiveness. We do note, however, that such data is naturally collected in a multimodal fashion, often paired with images. Rather than pre-training with only self-supervised objectives, we argue that it is better to bootstrap point cloud representations using image-based foundation models trained on internet-scale image data. Specifically, we propose a shelf-supervised approach (e.g. supervised with off-the-shelf image foundation models) for generating zero-shot 3D bounding boxes from paired RGB and LiDAR data. Pre-training 3D detectors with such pseudo-labels yields significantly better semi-supervised detection accuracy than prior self-supervised pretext tasks. Importantly, we show that image-based shelf-supervision is helpful for training LiDAR-only and multi-modal (RGB + LiDAR) detectors. We demonstrate the effectiveness of our approach on nuScenes and WOD, significantly improving over prior work in limited data settings. Our code is available at https://github.com/meharkhurana03/cm3d, Comment: The first two authors contributed equally. This work has been accepted to the Conference on Robot Learning (CoRL) 2024

Humans experience the world through constantly changing visual stimuli, where scenes can shift and move, change in appearance, and vary in distance. The dynamic nature of visual perception is a fundamental aspect of our daily lives, yet the large majority of research on object and scene processing, particularly using fMRI, has focused on static stimuli. While studies of static image perception are attractive due to their computational simplicity, they impose a strong non-naturalistic constraint on our investigation of human vision. In contrast, dynamic visual stimuli offer a more ecologically-valid approach but present new challenges due to the interplay between spatial and temporal information, making it difficult to disentangle the representations of stable image features and motion. To overcome this limitation -- given dynamic inputs, we explicitly decouple the modeling of static image representations and motion representations in the human brain. Three results demonstrate the feasibility of this approach. First, we show that visual motion information as optical flow can be predicted (or decoded) from brain activity as measured by fMRI. Second, we show that this predicted motion can be used to realistically animate static images using a motion-conditioned video diffusion model (where the motion is driven by fMRI brain activity). Third, we show prediction in the reverse direction: existing video encoders can be fine-tuned to predict fMRI brain activity from video imagery, and can do so more effectively than image encoders. This foundational work offers a novel, extensible framework for interpreting how the human brain processes dynamic visual information.

The integration of technology into exercise regimens has emerged as a strategy to enhance normal human capabilities and return human motor function after injury or illness by enhancing motor learning and retention. Much research has focused on how active devices, whether confined to a lab or made into a wearable format, can apply forces at set times and conditions to optimize the process of learning. However, the focus on active force production often forces devices to either be confined to simple movements or interventions. As such, in this paper, we investigate how passive device behaviors can contribute to the process of motor learning by themselves. Our approach involves using a wearable resistance (WR) device, which is outfitted with elastic bands, to apply a force field that changes in response to a person's movements while performing exercises. We develop a method to measure the produced forces from the device without impeding the function and we characterize the device's force generation abilities. We then present a study assessing the impact of the WR device on motor learning of proper squat form compared to visual or no feedback. Biometrics such as knee and hip angles were used to monitor and assess subject performance. Our findings indicate that the force fields produced while training with the WR device can improve performance in full-body exercises similarly to a more direct visual feedback mechanism, though the improvement is not consistent across all performance metrics. Through our research, we contribute important insights into the application of passive wearable resistance technology in practical exercise settings., Comment: 8 pages, 9 figures, To be published in IEEE International Conference on Biomedical Robotics and Biomechatronics (BioRob) 2024

Unsupervised pre-training has emerged as a transformative paradigm, displaying remarkable advancements in various domains. However, the susceptibility to domain shift, where pre-training data distribution differs from fine-tuning, poses a significant obstacle. To address this, we augment the Swin Transformer to learn from different medical imaging modalities, enhancing downstream performance. Our model, dubbed SwinFUSE (Swin Multi-Modal Fusion for UnSupervised Enhancement), offers three key advantages: (i) it learns from both Computed Tomography (CT) and Magnetic Resonance Images (MRI) during pre-training, resulting in complementary feature representations; (ii) a domain-invariance module (DIM) that effectively highlights salient input regions, enhancing adaptability; (iii) exhibits remarkable generalizability, surpassing the confines of tasks it was initially pre-trained on. Our experiments on two publicly available 3D segmentation datasets show a modest 1-2% performance trade-off compared to single-modality models, yet significant out-performance of up to 27% on out-of-distribution modality. This substantial improvement underscores our proposed approach's practical relevance and real-world applicability. Code is available at: https://github.com/devalab/SwinFUSE

Our work explores the task of generating future sensor observations conditioned on the past. We are motivated by `predictive coding' concepts from neuroscience as well as robotic applications such as self-driving vehicles. Predictive video modeling is challenging because the future may be multi-modal and learning at scale remains computationally expensive for video processing. To address both challenges, our key insight is to leverage the large-scale pretraining of image diffusion models which can handle multi-modality. We repurpose image models for video prediction by conditioning on new frame timestamps. Such models can be trained with videos of both static and dynamic scenes. To allow them to be trained with modestly-sized datasets, we introduce invariances by factoring out illumination and texture by forcing the model to predict (pseudo) depth, readily obtained for in-the-wild videos via off-the-shelf monocular depth networks. In fact, we show that simply modifying networks to predict grayscale pixels already improves the accuracy of video prediction. Given the extra controllability with timestamp conditioning, we propose sampling schedules that work better than the traditional autoregressive and hierarchical sampling strategies. Motivated by probabilistic metrics from the object forecasting literature, we create a benchmark for video prediction on a diverse set of videos spanning indoor and outdoor scenes and a large vocabulary of objects. Our experiments illustrate the effectiveness of learning to condition on timestamps, and show the importance of predicting the future with invariant modalities., Comment: Project page: http://www.cs.cmu.edu/~tkhurana/depthforecasting/

Despite significant progress in generative AI, comprehensive evaluation remains challenging because of the lack of effective metrics and standardized benchmarks. For instance, the widely-used CLIPScore measures the alignment between a (generated) image and text prompt, but it fails to produce reliable scores for complex prompts involving compositions of objects, attributes, and relations. One reason is that text encoders of CLIP can notoriously act as a "bag of words", conflating prompts such as "the horse is eating the grass" with "the grass is eating the horse". To address this, we introduce the VQAScore, which uses a visual-question-answering (VQA) model to produce an alignment score by computing the probability of a "Yes" answer to a simple "Does this figure show '{text}'?" question. Though simpler than prior art, VQAScore computed with off-the-shelf models produces state-of-the-art results across many (8) image-text alignment benchmarks. We also compute VQAScore with an in-house model that follows best practices in the literature. For example, we use a bidirectional image-question encoder that allows image embeddings to depend on the question being asked (and vice versa). Our in-house model, CLIP-FlanT5, outperforms even the strongest baselines that make use of the proprietary GPT-4V. Interestingly, although we train with only images, VQAScore can also align text with video and 3D models. VQAScore allows researchers to benchmark text-to-visual generation using complex texts that capture the compositional structure of real-world prompts. We introduce GenAI-Bench, a more challenging benchmark with 1,600 compositional text prompts that require parsing scenes, objects, attributes, relationships, and high-order reasoning like comparison and logic. GenAI-Bench also offers over 15,000 human ratings for leading image and video generation models such as Stable Diffusion, DALL-E 3, and Gen2., Comment: We open-source our data, model, and code at: https://github.com/linzhiqiu/t2v_metrics ; Project page: https://linzhiqiu.github.io/papers/vqascore

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 Lidar Panoptic Segmentation 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. Code and models are available at this $\href{https://github.com/nv-dvl/segment-anything-lidar}{URL}$., Comment: Accepted to ECCV 2024

Current scene flow methods broadly fail to describe motion on small objects, and current scene flow evaluation protocols hide this failure by averaging over many points, with most drawn larger objects. To fix this evaluation failure, we propose a new evaluation protocol, Bucket Normalized EPE, which is class-aware and speed-normalized, enabling contextualized error comparisons between object types that move at vastly different speeds. To highlight current method failures, we propose a frustratingly simple supervised scene flow baseline, TrackFlow, built by bolting a high-quality pretrained detector (trained using many class rebalancing techniques) onto a simple tracker, that produces state-of-the-art performance on current standard evaluations and large improvements over prior art on our new evaluation. Our results make it clear that all scene flow evaluations must be class and speed aware, and supervised scene flow methods must address point class imbalances. We release the evaluation code publicly at https://github.com/kylevedder/BucketedSceneFlowEval., Comment: Accepted to ECCV 2024. Project page at https://vedder.io/trackflow

Estimating camera poses is a fundamental task for 3D reconstruction and remains challenging given sparsely sampled views (<10). In contrast to existing approaches that pursue top-down prediction of global parametrizations of camera extrinsics, we propose a distributed representation of camera pose that treats a camera as a bundle of rays. This representation allows for a tight coupling with spatial image features improving pose precision. We observe that this representation is naturally suited for set-level transformers and develop a regression-based approach that maps image patches to corresponding rays. To capture the inherent uncertainties in sparse-view pose inference, we adapt this approach to learn a denoising diffusion model which allows us to sample plausible modes while improving performance. Our proposed methods, both regression- and diffusion-based, demonstrate state-of-the-art performance on camera pose estimation on CO3D while generalizing to unseen object categories and in-the-wild captures., Comment: In ICLR 2024 (oral). v2-3: updated references. Project webpage: https://jasonyzhang.com/RayDiffusion

Manually creating textures for 3D meshes is time-consuming, even for expert visual content creators. We propose a fast approach for automatically texturing an input 3D mesh based on a user-provided text prompt. Importantly, our approach disentangles lighting from surface material/reflectance in the resulting texture so that the mesh can be properly relit and rendered in any lighting environment. We introduce LightControlNet, a new text-to-image model based on the ControlNet architecture, which allows the specification of the desired lighting as a conditioning image to the model. Our text-to-texture pipeline then constructs the texture in two stages. The first stage produces a sparse set of visually consistent reference views of the mesh using LightControlNet. The second stage applies a texture optimization based on Score Distillation Sampling (SDS) that works with LightControlNet to increase the texture quality while disentangling surface material from lighting. Our algorithm is significantly faster than previous text-to-texture methods, while producing high-quality and relightable textures., Comment: Project page: https://flashtex.github.io/

Accurate and interpretable diagnostic models are crucial in the safety-critical field of medicine. We investigate the interpretability of our proposed biomarker-based lung ultrasound diagnostic pipeline to enhance clinicians' diagnostic capabilities. The objective of this study is to assess whether explanations from a decision tree classifier, utilizing biomarkers, can improve users' ability to identify inaccurate model predictions compared to conventional saliency maps. Our findings demonstrate that decision tree explanations, based on clinically established biomarkers, can assist clinicians in detecting false positives, thus improving the reliability of diagnostic models in medicine., Comment: Accepted at BIAS 2023 Conference

Vision-language models (VLMs) excel in zero-shot recognition but their performance varies greatly across different visual concepts. For example, although CLIP achieves impressive accuracy on ImageNet (60-80%), its performance drops below 10% for more than ten concepts like night snake, presumably due to their limited presence in the pretraining data. However, measuring the frequency of concepts in VLMs' large-scale datasets is challenging. We address this by using large language models (LLMs) to count the number of pretraining texts that contain synonyms of these concepts. Our analysis confirms that popular datasets, such as LAION, exhibit a long-tailed concept distribution, yielding biased performance in VLMs. We also find that downstream applications of VLMs, including visual chatbots (e.g., GPT-4V) and text-to-image models (e.g., Stable Diffusion), often fail to recognize or generate images of rare concepts identified by our method. To mitigate the imbalanced performance of zero-shot VLMs, we propose REtrieval-Augmented Learning (REAL). First, instead of prompting VLMs using the original class names, REAL uses their most frequent synonyms found in pretraining texts. This simple change already outperforms costly human-engineered and LLM-enriched prompts over nine benchmark datasets. Second, REAL trains a linear classifier on a small yet balanced set of pretraining data retrieved using concept synonyms. REAL surpasses the previous zero-shot SOTA, using 400x less storage and 10,000x less training time!, Comment: Project Page: https://shubhamprshr27.github.io/neglected-tails-of-vlms/

The era of vision-language models (VLMs) trained on large web-scale datasets challenges conventional formulations of "open-world" perception. In this work, we revisit the task of few-shot object detection (FSOD) in the context of recent foundational VLMs. First, we point out that zero-shot VLMs such as GroundingDINO significantly outperform state-of-the-art few-shot detectors (48 vs. 33 AP) on COCO. Despite their strong zero-shot performance, such foundational models may still be sub-optimal. For example, trucks on the web may be defined differently from trucks for a target application such as autonomous vehicle perception. We argue that the task of few-shot recognition can be reformulated as aligning foundation models to target concepts using a few examples. Interestingly, such examples can be multi-modal, using both text and visual cues, mimicking instructions that are often given to human annotators when defining a target concept of interest. Concretely, we propose Foundational FSOD, a new benchmark protocol that evaluates detectors pre-trained on any external datasets and fine-tuned on multi-modal (text and visual) K-shot examples per target class. We repurpose nuImages for Foundational FSOD, benchmark several popular open-source VLMs, and provide an empirical analysis of state-of-the-art methods. Lastly, we discuss our recent CVPR 2024 Foundational FSOD competition and share insights from the community. Notably, the winning team significantly outperforms our baseline by 23.9 mAP!

Amodal perception, the ability to comprehend complete object structures from partial visibility, is a fundamental skill, even for infants. Its significance extends to applications like autonomous driving, where a clear understanding of heavily occluded objects is essential. However, modern detection and tracking algorithms often overlook this critical capability, perhaps due to the prevalence of \textit{modal} annotations in most benchmarks. To address the scarcity of amodal benchmarks, we introduce TAO-Amodal, featuring 833 diverse categories in thousands of video sequences. Our dataset includes \textit{amodal} and modal bounding boxes for visible and partially or fully occluded objects, including those that are partially out of the camera frame. We investigate the current lay of the land in both amodal tracking and detection by benchmarking state-of-the-art modal trackers and amodal segmentation methods. We find that existing methods, even when adapted for amodal tracking, struggle to detect and track objects under heavy occlusion. To mitigate this, we explore simple finetuning schemes that can increase the amodal tracking and detection metrics of occluded objects by 2.1\% and 3.3\%., Comment: Project Page: https://tao-amodal.github.io

Long-Tailed 3D Object Detection (LT3D) addresses the problem of accurately detecting objects from both common and rare classes. Contemporary multi-modal detectors achieve low AP on rare-classes (e.g., CMT only achieves 9.4 AP on stroller), presumably because training detectors end-to-end with significant class imbalance is challenging. To address this limitation, we delve into a simple late-fusion framework that ensembles independently trained uni-modal LiDAR and RGB detectors. Importantly, such a late-fusion framework allows us to leverage large-scale uni-modal datasets (with more examples for rare classes) to train better uni-modal RGB detectors, unlike prevailing multimodal detectors that require paired multi-modal training data. Notably, our approach significantly improves rare-class detection by 7.2% over prior work. Further, we examine three critical components of our simple late-fusion approach from first principles and investigate whether to train 2D or 3D RGB detectors, whether to match RGB and LiDAR detections in 3D or the projected 2D image plane for fusion, and how to fuse matched detections. Extensive experiments reveal that 2D RGB detectors achieve better recognition accuracy for rare classes than 3D RGB detectors and matching on the 2D image plane mitigates depth estimation errors. Our late-fusion approach achieves 51.4 mAP on the established nuScenes LT3D benchmark, improving over prior work by 5.9 mAP!

Neural radiance fields provide state-of-the-art view synthesis quality but tend to be slow to render. One reason is that they make use of volume rendering, thus requiring many samples (and model queries) per ray at render time. Although this representation is flexible and easy to optimize, most real-world objects can be modeled more efficiently with surfaces instead of volumes, requiring far fewer samples per ray. This observation has spurred considerable progress in surface representations such as signed distance functions, but these may struggle to model semi-opaque and thin structures. We propose a method, HybridNeRF, that leverages the strengths of both representations by rendering most objects as surfaces while modeling the (typically) small fraction of challenging regions volumetrically. We evaluate HybridNeRF against the challenging Eyeful Tower dataset along with other commonly used view synthesis datasets. When comparing to state-of-the-art baselines, including recent rasterization-based approaches, we improve error rates by 15-30% while achieving real-time framerates (at least 36 FPS) for virtual-reality resolutions (2Kx2K)., Comment: CVPR 2024 Project page: https://haithemturki.com/hybrid-nerf/

Dense simultaneous localization and mapping (SLAM) is crucial for robotics and augmented reality applications. However, current methods are often hampered by the non-volumetric or implicit way they represent a scene. This work introduces SplaTAM, an approach that, for the first time, leverages explicit volumetric representations, i.e., 3D Gaussians, to enable high-fidelity reconstruction from a single unposed RGB-D camera, surpassing the capabilities of existing methods. SplaTAM employs a simple online tracking and mapping system tailored to the underlying Gaussian representation. It utilizes a silhouette mask to elegantly capture the presence of scene density. This combination enables several benefits over prior representations, including fast rendering and dense optimization, quickly determining if areas have been previously mapped, and structured map expansion by adding more Gaussians. Extensive experiments show that SplaTAM achieves up to 2x superior performance in camera pose estimation, map construction, and novel-view synthesis over existing methods, paving the way for more immersive high-fidelity SLAM applications., Comment: CVPR 2024. Website: https://spla-tam.github.io/

Neural Radiance Fields (NeRFs) can be dramatically accelerated by spatial grid representations. However, they do not explicitly reason about scale and so introduce aliasing artifacts when reconstructing scenes captured at different camera distances. Mip-NeRF and its extensions propose scale-aware renderers that project volumetric frustums rather than point samples but such approaches rely on positional encodings that are not readily compatible with grid methods. We propose a simple modification to grid-based models by training model heads at different spatial grid resolutions. At render time, we simply use coarser grids to render samples that cover larger volumes. Our method can be easily applied to existing accelerated NeRF methods and significantly improves rendering quality (reducing error rates by 20-90% across synthetic and unbounded real-world scenes) while incurring minimal performance overhead (as each model head is quick to evaluate). Compared to Mip-NeRF, we reduce error rates by 20% while training over 60x faster., Comment: Neurips 2023 Project page: https://haithemturki.com/pynerf/

Teaching robots to fold, drape, or reposition deformable objects such as cloth will unlock a variety of automation applications. While remarkable progress has been made for rigid object manipulation, manipulating deformable objects poses unique challenges, including frequent occlusions, infinite-dimensional state spaces and complex dynamics. Just as object pose estimation and tracking have aided robots for rigid manipulation, dense 3D tracking (scene flow) of highly deformable objects will enable new applications in robotics while aiding existing approaches, such as imitation learning or creating digital twins with real2sim transfer. We propose DeformGS, an approach to recover scene flow in highly deformable scenes, using simultaneous video captures of a dynamic scene from multiple cameras. DeformGS builds on recent advances in Gaussian splatting, a method that learns the properties of a large number of Gaussians for state-of-the-art and fast novel-view synthesis. DeformGS learns a deformation function to project a set of Gaussians with canonical properties into world space. The deformation function uses a neural-voxel encoding and a multilayer perceptron (MLP) to infer Gaussian position, rotation, and a shadow scalar. We enforce physics-inspired regularization terms based on conservation of momentum and isometry, which leads to trajectories with smaller trajectory errors. We also leverage existing foundation models SAM and XMEM to produce noisy masks, and learn a per-Gaussian mask for better physics-inspired regularization. DeformGS achieves high-quality 3D tracking on highly deformable scenes with shadows and occlusions. In experiments, DeformGS improves 3D tracking by an average of 55.8% compared to the state-of-the-art. With sufficient texture, DeformGS achieves a median tracking error of 3.3 mm on a cloth of 1.5 x 1.5 m in area. Website: https://deformgs.github.io

The rapid evolution of deep neural networks has revolutionized the field of machine learning, enabling remarkable advancements in various domains. In this article, we introduce NeuroWrite, a unique method for predicting the categorization of handwritten digits using deep neural networks. Our model exhibits outstanding accuracy in identifying and categorising handwritten digits by utilising the strength of convolutional neural networks (CNNs) and recurrent neural networks (RNNs).In this article, we give a thorough examination of the data preparation methods, network design, and training methods used in NeuroWrite. By implementing state-of-the-art techniques, we showcase how NeuroWrite can achieve high classification accuracy and robust generalization on handwritten digit datasets, such as MNIST. Furthermore, we explore the model's potential for real-world applications, including digit recognition in digitized documents, signature verification, and automated postal code recognition. NeuroWrite is a useful tool for computer vision and pattern recognition because of its performance and adaptability.The architecture, training procedure, and evaluation metrics of NeuroWrite are covered in detail in this study, illustrating how it can improve a number of applications that call for handwritten digit classification. The outcomes show that NeuroWrite is a promising method for raising the bar for deep neural network-based handwritten digit recognition., Comment: 6 pages, 10 figures

State-of-the-art lidar panoptic segmentation (LPS) methods follow bottom-up segmentation-centric fashion wherein they build upon semantic segmentation networks by utilizing clustering to obtain object instances. In this paper, we re-think this approach and propose a surprisingly simple yet effective detection-centric network for both LPS and tracking. Our network is modular by design and optimized for all aspects of both the panoptic segmentation and tracking task. One of the core components of our network is the object instance detection branch, which we train using point-level (modal) annotations, as available in segmentation-centric datasets. In the absence of amodal (cuboid) annotations, we regress modal centroids and object extent using trajectory-level supervision that provides information about object size, which cannot be inferred from single scans due to occlusions and the sparse nature of the lidar data. We obtain fine-grained instance segments by learning to associate lidar points with detected centroids. We evaluate our method on several 3D/4D LPS benchmarks and observe that our model establishes a new state-of-the-art among open-sourced models, outperforming recent query-based models., Comment: IROS 2023. Code at https://github.com/abhinavagarwalla/most-lps

Autonomous vehicles (AVs) require comprehensive and reliable pedestrian trajectory data to ensure safe operation. However, obtaining data of safety-critical scenarios such as jaywalking and near-collisions, or uncommon agents such as children, disabled pedestrians, and vulnerable road users poses logistical and ethical challenges. This paper evaluates a Virtual Reality (VR) system designed to collect pedestrian trajectory and body pose data in a controlled, low-risk environment. We substantiate the usefulness of such a system through semi-structured interviews with professionals in the AV field, and validate the effectiveness of the system through two empirical studies: a first-person user evaluation involving 62 participants, and a third-person evaluative survey involving 290 respondents. Our findings demonstrate that the VR-based data collection system elicits realistic responses for capturing pedestrian data in safety-critical or uncommon vehicle-pedestrian interaction scenarios., Comment: Spotlight paper in the Data Generation for Robotics Workshop at RSS 2024

Trajectory forecasting is a widely-studied problem for autonomous navigation. However, existing benchmarks evaluate forecasting based on independent snapshots of trajectories, which are not representative of real-world applications that operate on a continuous stream of data. To bridge this gap, we introduce a benchmark that continuously queries future trajectories on streaming data and we refer to it as "streaming forecasting." Our benchmark inherently captures the disappearance and re-appearance of agents, presenting the emergent challenge of forecasting for occluded agents, which is a safety-critical problem yet overlooked by snapshot-based benchmarks. Moreover, forecasting in the context of continuous timestamps naturally asks for temporal coherence between predictions from adjacent timestamps. Based on this benchmark, we further provide solutions and analysis for streaming forecasting. We propose a plug-and-play meta-algorithm called "Predictive Streamer" that can adapt any snapshot-based forecaster into a streaming forecaster. Our algorithm estimates the states of occluded agents by propagating their positions with multi-modal trajectories, and leverages differentiable filters to ensure temporal consistency. Both occlusion reasoning and temporal coherence strategies significantly improve forecasting quality, resulting in 25% smaller endpoint errors for occluded agents and 10-20% smaller fluctuations of trajectories. Our work is intended to generate interest within the community by highlighting the importance of addressing motion forecasting in its intrinsic streaming setting. Code is available at https://github.com/ziqipang/StreamingForecasting., Comment: IROS 2023, 8 pages, 9 figures

Vision-language models (VLMs) pre-trained on web-scale datasets have demonstrated remarkable capabilities on downstream tasks when fine-tuned with minimal data. However, many VLMs rely on proprietary data and are not open-source, which restricts the use of white-box approaches for fine-tuning. As such, we aim to develop a black-box approach to optimize VLMs through natural language prompts, thereby avoiding the need to access model parameters, feature embeddings, or even output logits. We propose employing chat-based LLMs to search for the best text prompt for VLMs. Specifically, we adopt an automatic hill-climbing procedure that converges to an effective prompt by evaluating the performance of current prompts and asking LLMs to refine them based on textual feedback, all within a conversational process without human-in-the-loop. In a challenging 1-shot image classification setup, our simple approach surpasses the white-box continuous prompting method (CoOp) by an average of 1.5% across 11 datasets including ImageNet. Our approach also outperforms both human-engineered and LLM-generated prompts. We highlight the advantage of conversational feedback that incorporates both positive and negative prompts, suggesting that LLMs can utilize the implicit gradient direction in textual feedback for a more efficient search. In addition, we find that the text prompts generated through our strategy are not only more interpretable but also transfer well across different VLM architectures in a black-box manner. Lastly, we apply our framework to optimize the state-of-the-art black-box VLM (DALL-E 3) for text-to-image generation, prompt inversion, and personalization., Comment: Published at CVPR 2024. Project site: https://llm-can-optimize-vlm.github.io/

We present a method that simultaneously addresses the tasks of dynamic scene novel-view synthesis and six degree-of-freedom (6-DOF) tracking of all dense scene elements. We follow an analysis-by-synthesis framework, inspired by recent work that models scenes as a collection of 3D Gaussians which are optimized to reconstruct input images via differentiable rendering. To model dynamic scenes, we allow Gaussians to move and rotate over time while enforcing that they have persistent color, opacity, and size. By regularizing Gaussians' motion and rotation with local-rigidity constraints, we show that our Dynamic 3D Gaussians correctly model the same area of physical space over time, including the rotation of that space. Dense 6-DOF tracking and dynamic reconstruction emerges naturally from persistent dynamic view synthesis, without requiring any correspondence or flow as input. We demonstrate a large number of downstream applications enabled by our representation, including first-person view synthesis, dynamic compositional scene synthesis, and 4D video editing.

Resource-constrained perception systems such as edge computing and vision-for-robotics require vision models to be both accurate and lightweight in computation and memory usage. While knowledge distillation is a proven strategy to enhance the performance of lightweight classification models, its application to structured outputs like object detection and instance segmentation remains a complicated task, due to the variability in outputs and complex internal network modules involved in the distillation process. In this paper, we propose a simple yet surprisingly effective sequential approach to knowledge distillation that progressively transfers the knowledge of a set of teacher detectors to a given lightweight student. To distill knowledge from a highly accurate but complex teacher model, we construct a sequence of teachers to help the student gradually adapt. Our progressive strategy can be easily combined with existing detection distillation mechanisms to consistently maximize student performance in various settings. To the best of our knowledge, we are the first to successfully distill knowledge from Transformer-based teacher detectors to convolution-based students, and unprecedentedly boost the performance of ResNet-50 based RetinaNet from 36.5% to 42.0% AP and Mask R-CNN from 38.2% to 42.5% AP on the MS COCO benchmark., Comment: ICML 2023

Computing platforms in autonomous vehicles record large amounts of data from many sensors, process the data through machine learning models, and make decisions to ensure the vehicle's safe operation. Fast, accurate, and reliable decision-making is critical. Traditional computer processors lack the power and flexibility needed for the perception and machine vision demands of advanced autonomous driving tasks. Hardware accelerators are special-purpose coprocessors that help autonomous vehicles meet performance requirements for higher levels of autonomy. This paper provides an overview of ML accelerators with examples of their use for machine vision in autonomous vehicles. We offer recommendations for researchers and practitioners and highlight a trajectory for ongoing and future research in this emerging field.

LiDAR-based 3D detection plays a vital role in autonomous navigation. Surprisingly, although autonomous vehicles (AVs) must detect both near-field objects (for collision avoidance) and far-field objects (for longer-term planning), contemporary benchmarks focus only on near-field 3D detection. However, AVs must detect far-field objects for safe navigation. In this paper, we present an empirical analysis of far-field 3D detection using the long-range detection dataset Argoverse 2.0 to better understand the problem, and share the following insight: near-field LiDAR measurements are dense and optimally encoded by small voxels, while far-field measurements are sparse and are better encoded with large voxels. We exploit this observation to build a collection of range experts tuned for near-vs-far field detection, and propose simple techniques to efficiently ensemble models for long-range detection that improve efficiency by 33% and boost accuracy by 3.2% CDS., Comment: Accepted to ICCV 2023 Workshop - Robustness and Reliability of Autonomous Vehicles in the Open-World

Large-scale datasets are essential to modern day deep learning. Advocates argue that understanding these methods requires dataset transparency (e.g. "dataset curation, motivation, composition, collection process, etc..."). However, almost no one has suggested the release of the detailed definitions and visual category examples provided to annotators - information critical to understanding the structure of the annotations present in each dataset. These labels are at the heart of public datasets, yet few datasets include the instructions that were used to generate them. We introduce a new task, Labeling Instruction Generation, to address missing publicly available labeling instructions. In Labeling Instruction Generation, we take a reasonably annotated dataset and: 1) generate a set of examples that are visually representative of each category in the dataset; 2) provide a text label that corresponds to each of the examples. We introduce a framework that requires no model training to solve this task and includes a newly created rapid retrieval system that leverages a large, pre-trained vision and language model. This framework acts as a proxy to human annotators that can help to both generate a final labeling instruction set and evaluate its quality. Our framework generates multiple diverse visual and text representations of dataset categories. The optimized instruction set outperforms our strongest baseline across 5 folds by 7.06 mAP for NuImages and 12.9 mAP for COCO.