We introduce a new benchmark, TAPVid-3D, for evaluating the task of long-range Tracking Any Point in 3D (TAP-3D). While point tracking in two dimensions (TAP) has many benchmarks measuring performance on real-world videos, such as TAPVid-DAVIS, three-dimensional point tracking has none. To this end, leveraging existing footage, we build a new benchmark for 3D point tracking featuring 4,000+ real-world videos, composed of three different data sources spanning a variety of object types, motion patterns, and indoor and outdoor environments. To measure performance on the TAP-3D task, we formulate a collection of metrics that extend the Jaccard-based metric used in TAP to handle the complexities of ambiguous depth scales across models, occlusions, and multi-track spatio-temporal smoothness. We manually verify a large sample of trajectories to ensure correct video annotations, and assess the current state of the TAP-3D task by constructing competitive baselines using existing tracking models. We anticipate this benchmark will serve as a guidepost to improve our ability to understand precise 3D motion and surface deformation from monocular video. Code for dataset download, generation, and model evaluation is available at https://tapvid3d.github.io
The Video comprises approximately ~85\% of all internet traffic, but video encoding/compression is being historically done with hard coded rules, which has worked well but only to a certain limit. We have seen a surge in video compression algorithms using ML-based models in the last few years and many of them have outperformed several legacy codecs. The models range from encoding video end to end using an ML approach or replacing some intermediate steps in legacy codecs using ML models to increase the efficiency of those steps. Optimizing video storage is an essential aspect of video processing, so we are proposing one of the possible approaches to achieve it is by avoiding redundant data at each frame. In this paper, we want to introduce the approach of redundancies removal in subsequent frames for a given video as a main approach for video compression. We call this method Redundancy Removal using Shift (R\textsuperscript2S). This method can be utilized across various Machine Learning model algorithms, and make the compression more accessible and adaptable. In this study, we have utilized a computer vision-based pixel point tracking method to identify redundant pixels to encode video for optimal storage.
Gaussian splatting has become a popular representation for novel-view synthesis, exhibiting clear strengths in efficiency, photometric quality, and compositional edibility. Following its success, many works have extended Gaussians to 4D, showing that dynamic Gaussians maintain these benefits while also tracking scene geometry far better than alternative representations. Yet, these methods assume dense multi-view videos as supervision. In this work, we are interested in extending the capability of Gaussian scene representations to casually captured monocular videos. We show that existing 4D Gaussian methods dramatically fail in this setup because the monocular setting is underconstrained. Building off this finding, we propose a method we call Dynamic Gaussian Marbles, which consist of three core modifications that target the difficulties of the monocular setting. First, we use isotropic Gaussian "marbles'', reducing the degrees of freedom of each Gaussian. Second, we employ a hierarchical divide and-conquer learning strategy to efficiently guide the optimization towards solutions with globally coherent motion. Finally, we add image-level and geometry-level priors into the optimization, including a tracking loss that takes advantage of recent progress in point tracking. By constraining the optimization, Dynamic Gaussian Marbles learns Gaussian trajectories that enable novel-view rendering and accurately capture the 3D motion of the scene elements. We evaluate on the Nvidia Dynamic Scenes dataset and the DyCheck iPhone dataset, and show that Gaussian Marbles significantly outperforms other Gaussian baselines in quality, and is on-par with non-Gaussian representations, all while maintaining the efficiency, compositionality, editability, and tracking benefits of Gaussians. Our project page can be found here https://geometry.stanford.edu/projects/dynamic-gaussian-marbles.github.io/.
Interactive video object segmentation is a crucial video task, having various applications from video editing to data annotating. However, current approaches struggle to accurately segment objects across diverse domains. Recently, Segment Anything Model (SAM) introduces interactive visual prompts and demonstrates impressive performance across different domains. In this paper, we propose a training-free prompt tracking framework for interactive video object segmentation (I-PT), leveraging the powerful generalization of SAM. Although point tracking efficiently captures the pixel-wise information of objects in a video, points tend to be unstable when tracked over a long period, resulting in incorrect segmentation. Towards fast and robust interaction, we jointly adopt sparse points and boxes tracking, filtering out unstable points and capturing object-wise information. To better integrate reference information from multiple interactions, we introduce a cross-round space-time module (CRSTM), which adaptively aggregates mask features from previous rounds and frames, enhancing the segmentation stability. Our framework has demonstrated robust zero-shot video segmentation results on popular VOS datasets with interaction types, including DAVIS 2017, YouTube-VOS 2018, and MOSE 2023, maintaining a good tradeoff between performance and interaction time.
Tissue tracking in echocardiography is challenging due to the complex cardiac motion and the inherent nature of ultrasound acquisitions. Although optical flow methods are considered state-of-the-art (SOTA), they struggle with long-range tracking, noise occlusions, and drift throughout the cardiac cycle. Recently, novel learning-based point tracking techniques have been introduced to tackle some of these issues. In this paper, we build upon these techniques and introduce EchoTracker, a two-fold coarse-to-fine model that facilitates the tracking of queried points on a tissue surface across ultrasound image sequences. The architecture contains a preliminary coarse initialization of the trajectories, followed by reinforcement iterations based on fine-grained appearance changes. It is efficient, light, and can run on mid-range GPUs. Experiments demonstrate that the model outperforms SOTA methods, with an average position accuracy of 67% and a median trajectory error of 2.86 pixels. Furthermore, we show a relative improvement of 25% when using our model to calculate the global longitudinal strain (GLS) in a clinical test-retest dataset compared to other methods. This implies that learning-based point tracking can potentially improve performance and yield a higher diagnostic and prognostic value for clinical measurements than current techniques. Our source code is available at: https://github.com/riponazad/echotracker/.
This paper introduces FlowMap, an end-to-end differentiable method that solves for precise camera poses, camera intrinsics, and per-frame dense depth of a video sequence. Our method performs per-video gradient-descent minimization of a simple least-squares objective that compares the optical flow induced by depth, intrinsics, and poses against correspondences obtained via off-the-shelf optical flow and point tracking. Alongside the use of point tracks to encourage long-term geometric consistency, we introduce differentiable re-parameterizations of depth, intrinsics, and pose that are amenable to first-order optimization. We empirically show that camera parameters and dense depth recovered by our method enable photo-realistic novel view synthesis on 360-degree trajectories using Gaussian Splatting. Our method not only far outperforms prior gradient-descent based bundle adjustment methods, but surprisingly performs on par with COLMAP, the state-of-the-art SfM method, on the downstream task of 360-degree novel view synthesis (even though our method is purely gradient-descent based, fully differentiable, and presents a complete departure from conventional SfM).
This paper addresses the long-standing challenge of reconstructing 3D structures from videos with dynamic content. Current approaches to this problem were not designed to operate on casual videos recorded by standard cameras or require a long optimization time. Aiming to significantly improve the efficiency of previous approaches, we present TracksTo4D, a learning-based approach that enables inferring 3D structure and camera positions from dynamic content originating from casual videos using a single efficient feed-forward pass. To achieve this, we propose operating directly over 2D point tracks as input and designing an architecture tailored for processing 2D point tracks. Our proposed architecture is designed with two key principles in mind: (1) it takes into account the inherent symmetries present in the input point tracks data, and (2) it assumes that the movement patterns can be effectively represented using a low-rank approximation. TracksTo4D is trained in an unsupervised way on a dataset of casual videos utilizing only the 2D point tracks extracted from the videos, without any 3D supervision. Our experiments show that TracksTo4D can reconstruct a temporal point cloud and camera positions of the underlying video with accuracy comparable to state-of-the-art methods, while drastically reducing runtime by up to 95\%. We further show that TracksTo4D generalizes well to unseen videos of unseen semantic categories at inference time.
This report proposes an improved method for the Tracking Any Point (TAP) task, which tracks any physical surface through a video. Several existing approaches have explored the TAP by considering the temporal relationships to obtain smooth point motion trajectories, however, they still suffer from the cumulative error caused by temporal prediction. To address this issue, we propose a simple yet effective approach called TAP with confident static points (TAPIR+), which focuses on rectifying the tracking of the static point in the videos shot by a static camera. To clarify, our approach contains two key components: (1) Multi-granularity Camera Motion Detection, which could identify the video sequence by the static camera shot. (2) CMR-based point trajectory prediction with one moving object segmentation approach to isolate the static point from the moving object. Our approach ranked first in the final test with a score of 0.46.
We present DINO-Tracker -- a new framework for long-term dense tracking in video. The pillar of our approach is combining test-time training on a single video, with the powerful localized semantic features learned by a pre-trained DINO-ViT model. Specifically, our framework simultaneously adopts DINO's features to fit to the motion observations of the test video, while training a tracker that directly leverages the refined features. The entire framework is trained end-to-end using a combination of self-supervised losses, and regularization that allows us to retain and benefit from DINO's semantic prior. Extensive evaluation demonstrates that our method achieves state-of-the-art results on known benchmarks. DINO-tracker significantly outperforms self-supervised methods and is competitive with state-of-the-art supervised trackers, while outperforming them in challenging cases of tracking under long-term occlusions.
In this paper, we propose a simple and strong framework for Tracking Any Point with TRansformers (TAPTR). Based on the observation that point tracking bears a great resemblance to object detection and tracking, we borrow designs from DETR-like algorithms to address the task of TAP. In the proposed framework, in each video frame, each tracking point is represented as a point query, which consists of a positional part and a content part. As in DETR, each query (its position and content feature) is naturally updated layer by layer. Its visibility is predicted by its updated content feature. Queries belonging to the same tracking point can exchange information through self-attention along the temporal dimension. As all such operations are well-designed in DETR-like algorithms, the model is conceptually very simple. We also adopt some useful designs such as cost volume from optical flow models and develop simple designs to provide long temporal information while mitigating the feature drifting issue. Our framework demonstrates strong performance with state-of-the-art performance on various TAP datasets with faster inference speed.
Utilizing distribution grid flexibility for ancillary services requires the coordination and dispatch of requested active and reactive power to a large number of distributed energy resources in underlying grid layers. This paper presents an approach to hierarchically dispatch flexibility requests based on Online Feedback Optimization (OFO). We implement a framework of individual controllers coordinating actors, contributing to flexibility provision, to track a requested operating point at the interface between grid layers. The framework is evaluated in terms of performance during coordination and possible interaction between individual controllers, both central and distributed. Results show high reliability and robustness of the OFO controllers as well as an efficient dispatch of active and reactive power. Its computational efficiency and capabilities in set point tracking during online grid operation are making OFO a promising approach to the flexibility dispatch problem.
The Segment Anything Model (SAM) is a powerful vision foundation model that is revolutionizing the traditional paradigm of segmentation. Despite this, a reliance on prompting each frame and large computational cost limit its usage in robotically assisted surgery. Applications, such as augmented reality guidance, require little user intervention along with efficient inference to be usable clinically. In this study, we address these limitations by adopting lightweight SAM variants to meet the efficiency requirement and employing fine-tuning techniques to enhance their generalization in surgical scenes. Recent advancements in Tracking Any Point (TAP) have shown promising results in both accuracy and efficiency, particularly when points are occluded or leave the field of view. Inspired by this progress, we present a novel framework that combines an online point tracker with a lightweight SAM model that is fine-tuned for surgical instrument segmentation. Sparse points within the region of interest are tracked and used to prompt SAM throughout the video sequence, providing temporal consistency. The quantitative results surpass the state-of-the-art semi-supervised video object segmentation method XMem on the EndoVis 2015 dataset with 84.8 IoU and 91.0 Dice. Our method achieves promising performance that is comparable to XMem and transformer-based fully supervised segmentation methods on ex vivo UCL dVRK and in vivo CholecSeg8k datasets. In addition, the proposed method shows promising zero-shot generalization ability on the label-free STIR dataset. In terms of efficiency, we tested our method on a single GeForce RTX 4060/4090 GPU respectively, achieving an over 25/90 FPS inference speed. Code is available at: https://github.com/wuzijian1997/SIS-PT-SAM
Finding point-level correspondences is a fundamental problem in ultrasound (US), since it can enable US landmark tracking for intraoperative image guidance in different surgeries, including head and neck. Most existing US tracking methods, e.g., those based on optical flow or feature matching, were initially designed for RGB images before being applied to US. Therefore domain shift can impact their performance. Training could be supervised by ground-truth correspondences, but these are expensive to acquire in US. To solve these problems, we propose a self-supervised pixel-level tracking model called PIPsUS. Our model can track an arbitrary number of points in one forward pass and exploits temporal information by considering multiple, instead of just consecutive, frames. We developed a new self-supervised training strategy that utilizes a long-term point-tracking model trained for RGB images as a teacher to guide the model to learn realistic motions and use data augmentation to enforce tracking from US appearance. We evaluate our method on neck and oral US and echocardiography, showing higher point tracking accuracy when compared with fast normalized cross-correlation and tuned optical flow. Code will be available once the paper is accepted.
Point-based image editing has attracted remarkable attention since the emergence of DragGAN. Recently, DragDiffusion further pushes forward the generative quality via adapting this dragging technique to diffusion models. Despite these great success, this dragging scheme exhibits two major drawbacks, namely inaccurate point tracking and incomplete motion supervision, which may result in unsatisfactory dragging outcomes. To tackle these issues, we build a stable and precise drag-based editing framework, coined as StableDrag, by designing a discirminative point tracking method and a confidence-based latent enhancement strategy for motion supervision. The former allows us to precisely locate the updated handle points, thereby boosting the stability of long-range manipulation, while the latter is responsible for guaranteeing the optimized latent as high-quality as possible across all the manipulation steps. Thanks to these unique designs, we instantiate two types of image editing models including StableDrag-GAN and StableDrag-Diff, which attains more stable dragging performance, through extensive qualitative experiments and quantitative assessment on DragBench.
In contrast to set-point tracking which aims to reduce the tracking error between the tracker and the reference, tracking-in-range problems only focus on whether the tracker is within a given range around the reference, making it more suitable for the mission specifications of many practical applications. In this work, we present novel optimal control formulations to solve tracking-in-range problems, for both problems requiring the tracker to be always in range, and problems allowing the tracker to go out of range to yield overall better outcomes. As the problem naturally involves discontinuous functions, we present alternative formulations and regularisation strategies to improve the performance of numerical solvers. The extension to in-range tracking with multiple trackers and in-range tracking in high dimensional space are also discussed and illustrated with numerical examples, demonstrating substantial increases in mission duration in comparison to traditional set-point tracking.
Full-body avatar presence is crucial for immersive social and environmental interactions in digital reality. However, current devices only provide three six degrees of freedom (DOF) poses from the headset and two controllers (i.e. three-point trackers). Because it is a highly under-constrained problem, inferring full-body pose from these inputs is challenging, especially when supporting the full range of body proportions and use cases represented by the general population. In this paper, we propose a deep learning framework, DivaTrack, which outperforms existing methods when applied to diverse body sizes and activities. We augment the sparse three-point inputs with linear accelerations from Inertial Measurement Units (IMU) to improve foot contact prediction. We then condition the otherwise ambiguous lower-body pose with the predictions of foot contact and upper-body pose in a two-stage model. We further stabilize the inferred full-body pose in a wide range of configurations by learning to blend predictions that are computed in two reference frames, each of which is designed for different types of motions. We demonstrate the effectiveness of our design on a large dataset that captures 22 subjects performing challenging locomotion for three-point tracking, including lunges, hula-hooping, and sitting. As shown in a live demo using the Meta VR headset and Xsens IMUs, our method runs in real-time while accurately tracking a user's motion when they perform a diverse set of movements.
To endow models with greater understanding of physics and motion, it is useful to enable them to perceive how solid surfaces move and deform in real scenes. This can be formalized as Tracking-Any-Point (TAP), which requires the algorithm to track any point on solid surfaces in a video, potentially densely in space and time. Large-scale groundtruth training data for TAP is only available in simulation, which currently has a limited variety of objects and motion. In this work, we demonstrate how large-scale, unlabeled, uncurated real-world data can improve a TAP model with minimal architectural changes, using a selfsupervised student-teacher setup. We demonstrate state-of-the-art performance on the TAP-Vid benchmark surpassing previous results by a wide margin: for example, TAP-Vid-DAVIS performance improves from 61.3% to 67.4%, and TAP-Vid-Kinetics from 57.2% to 62.5%. For visualizations, see our project webpage at https://bootstap.github.io/
Grid-tied photovoltaic (PV) installations with Distributed Maximum Power Point Tracking (DMPPT) architectures include a DC-DC Module Integrated Converter (MIC) for managing each PV panel, isolating it from the others, reducing the mismatching effect and maximizing the harvested power. In this paper, the Autotransformer Forward converter with type-Zeta resonant reset (AFZ) is proposed as a DMPPT architecture MIC candidate. The main characteristics of the AFZ converter are the high versatility due to its voltage step-up and step-down capability; the use of an optimized autotransformer with only two windings, reducing the complexity and power losses of this component; the good dynamic performances, like the Forward converter ones; the low number of components and the simplicity and high feasibility associated to the use of just one active switch. Besides, soft switching transitions are achieved thanks to the autotransformer type-Zeta resonant reset. The steady-state theoretical analysis, considering the effect of the autotransformer leakage inductance, is presented. The converter is also studied in the frequency domain, obtaining the small-signal transfer functions. A design procedure based on the requirements of a 100 kW grid-tied photovoltaic installation is described, yielding in a 225 W prototype with efficiencies up to 95.6 %. Experimental results validate the theoretical analysis.
The mismatching phenomenon is one of the main issues in photovoltaic (PV) applications. It could reduce the generated power of a string when a PV panel has different performances from the other PV panels connected to the same string. Distributed Maximum Power Point Tracking (DMPPT) architectures are one of the most promising solutions to overcome the drawbacks associated with mismatching phenomena in PV applications. In this kind of architectures, a DC-DC module integrated converter (MIC) manages each PV panel, isolating it from the rest of the PV panels, for harvesting the maximum available power from the Sun. Due to the high number of DCDC converters used in a grid-tied PV installation, the most desired MIC requirements are high efficiency, low cost and the capability of voltage step-up and step-down. This paper proposes the Buck-Boost Modified Forward (BBMSF) converter as a good candidate to be applied in DMPPT architectures. A complete analysis of the BBMSF converter is carried out, including the steady-state analysis as well as the small signal analysis in continuous conduction mode. The main advantages of the BBMSF converter are its step-up and step-down voltage transfer function; a higher simplicity, since it only includes a single controlled switch; the soft switching characteristics in all the diodes and MOSFET, reaching in some cases ZVS and ZCS, and yielding high efficiencies; the use of an autotransformer, with better performances than a typical Forward transformer; and the good dynamic performance, like the Forward converter ones. The theoretical analyses are validated through the experimental results in a 225 W BBMSF prototype designed and built under the requirements of a 100 kW grid-tied PV installation, achieving an efficiency up to 93.6%.
Downstream wind turbines operating behind upstream turbines face significant performance challenges due to reduced wind speeds and increased turbulence. This leads to decreased wind energy production and higher dynamic loads on downwind turbines. Consequently, real-time monitoring and control have become crucial for improving wind farm performance. One promising solution involves optimizing wind farm layouts in real-time, taking advantage of the added flexibility offered by floating offshore wind turbines (FOWTs). This study explores a dynamic layout optimization strategy to minimize wake effects in wind farms while meeting power requirements. Two scenarios are considered: power maximization and power set-point tracking. The methodology involves a centralized wind farm controller optimizing the layout, followed by wind turbine controllers to meet the prescribed targets. Each FOWT employs model predictive control to adjust aerodynamic thrust force. The control strategy integrates a dynamic wind farm model that considers floating platform motion and wake transport in changing wind conditions. In a case study with a 1x3 wind farm layout of 5 MW FOWTs, the results show a 25% increase in stable energy production compared to a static layout in one hour for the first scenario. In the second scenario, desired power production was swiftly and consistently achieved.