arXiv subjects

Yihong Sun

Publications and source records attributed to Yihong Sun.

15 recordsLinked to original sources

arXiv2026

Unified Video Dense Prediction from Disjoint Data

Scene understanding requires simultaneous prediction about geometry, appearance, and semantics. However, existing task-specific annotations are fragmented across incompatible, domain-specific datasets. Current unified systems circumvent this by restricting training to fully co-annotated data, or by incurring the large computational cost of pseudo-labeling. To mitigate this, we introduce UniD, a unified video model that jointly predicts eight dense scene properties-depth, surface normals, semantic segmentation, boundaries, human parts, albedo, shading, and materials-all learned from disjoint, domain-specific datasets. We propose a simple yet effective distillation step in which per-task experts supervise a unified backbone through lightweight task projectors, eliminating the need for annotation overlap or pseudo-labeling. Our key insight is that the strong visual priors of a pretrained diffusion model are sufficient to bridge the domain gaps introduced by disjoint training sources, enabling robust generalization to scene-task combinations never seen during training. UniD achieves competitive performance against per-task specialists and multi-task baselines, with strong generalization to out-of-distribution scenarios and enhanced temporal and cross-task consistency. Code and video results are available at https://unid-video.github.io/.

cs.CV↗

arXiv2026

Efficient Tracking and Understanding Object Transformations

Tracking objects through state transformations is essential for understanding real-world dynamics. However, existing methods are computationally expensive. TubeletGraph recently showed impressive capabilities, but its inference cost (~$4.4$ seconds per object-frame on VOST) precludes any real-time deployment possibilities. We observe that TubeletGraph's overhead arises from building a spatiotemporal partition of the input video: (1) entity segmentation is computed densely for every frame regardless of whether a transformation occurs, and (2) every entity in the scene is tracked, scaling cost with scene complexity rather than the number of transformations of interest. To address both, we propose FluxGraph, a reactive variant that uses SAM2's internal multi-mask disagreement as a lightweight trigger for transformation detection, and removes the need for tracking all entities in the given video. FluxGraph is ~$3.3\times$ faster than TubeletGraph on VOST while improving tracking performance and preserving state graph quality. Furthermore, we also observe consistent speedups of $3.7-10.7\times$ across VSCOS, M$^3$-VOS, and DAVIS17 while maintaining performance. Code is publicly available at https://github.com/YihongSun/FluxGraph.

cs.CV↗

arXiv2026

Live Interactive Training for Video Segmentation

Interactive video segmentation often requires many user interventions for robust performance in challenging scenarios (e.g., occlusions, object separations, camouflage, etc.). Yet, even state-of-the-art models like SAM2 use corrections only for immediate fixes without learning from this feedback, leading to inefficient, repetitive user effort. To address this, we introduce Live Interactive Training (LIT), a novel framework for prompt-based visual systems where models also learn online from human corrections at inference time. Our primary instantiation, LIT-LoRA, implements this by continually updating a lightweight LoRA module on-the-fly. When a user provides a correction, this module is rapidly trained on that feedback, allowing the vision system to improve performance on subsequent frames of the same video. Leveraging the core principles of LIT, our LIT-LoRA implementation achieves an average 18-34% reduction in total corrections on challenging video segmentation benchmarks, with a negligible training overhead of ~0.5s per correction. We further demonstrate its generality by successfully adapting it to other segmentation models and extending it to CLIP-based fine-grained image classification. Our work highlights the promise of live adaptation to transform interactive tools and significantly reduce redundant human effort in complex visual tasks. Project: https://youngxinyu1802.github.io/projects/LIT/.

cs.CV↗

arXiv2025

Tracking and Understanding Object Transformations

Real-world objects frequently undergo state transformations. From an apple being cut into pieces to a butterfly emerging from its cocoon, tracking through these changes is important for understanding real-world objects and dynamics. However, existing methods often lose track of the target object after transformation, due to significant changes in object appearance. To address this limitation, we introduce the task of Track Any State: tracking objects through transformations while detecting and describing state changes, accompanied by a new benchmark dataset, VOST-TAS. To tackle this problem, we present TubeletGraph, a zero-shot system that recovers missing objects after transformation and maps out how object states are evolving over time. TubeletGraph first identifies potentially overlooked tracks, and determines whether they should be integrated based on semantic and proximity priors. Then, it reasons about the added tracks and generates a state graph describing each observed transformation. TubeletGraph achieves state-of-the-art tracking performance under transformations, while demonstrating deeper understanding of object transformations and promising capabilities in temporal grounding and semantic reasoning for complex object transformations. Code, additional results, and the benchmark dataset are available at https://tubelet-graph.github.io.

cs.CV↗

arXiv2025

Contactless Modulation of Intralayer and Interlayer Excitons in MoS2/WSe2 heterostructures with Acoustoelectric Fields

This work presents a platform that enables surface acoustic wave (SAW) modulation of both intralayer and interlayer excitons in MoS2/WSe2 heterostructures. Harnessing the coupled piezoelectric and strain fields of SAWs, this integrated approach allows for dynamic, precise, and fully contactless control of excitonic properties, a capability essential for the realization of next generation optoelectronic, quantum photonic, and excitonic devices. We identify two distinct modulable interlayer excitons in optical communication bands: IX$_{K\Gamma}$ in the O band (around 1300 nm) and IX$_{K\!-\!K}$ in the S band (around 1500 nm); these two excitons display a robust twist-angle-independent energy splitting of 120 meV, in agreement with density functional theory (DFT) calculations. The type-II band alignment induced by the SAW not only promotes efficient exciton dissociation but also enables direct and tunable modulation of photoluminescence via the formation of confined piezoelectric potential wells. Furthermore, by simultaneously generating in-plane and out-of-plane SAW fields, the platform achieves selective manipulation of intralayer and interlayer excitons, inducing quadratic Stark effects for intralayer excitons and linear Stark effects for interlayer excitons. These findings provide new insights into SAWexciton interactions in van der Waals heterostructures, broaden the operational spectral range, and establish pathways toward on-chip acousto-optic and quantum optoelectronic devices with advanced excitonic functionality.

physics.optics↗

arXiv2025

Comprehensive Analog Signal Processing Platform Enabled with Acoustic Charge Transport in Two-dimensional Materials

Two-dimensional Acoustic Charge Transport (2D-ACT) devices, which integrate two dimensional semiconductor field-effect transistor (FET) with high-frequency surface acoustic wave (SAW) device provide a potential compact platform for the processing of analog signals in a wireless, non-contact, low-loss and real-time way. It is expected to be used in long-distance space communication and sensing. However, current investigations into 2D-ACT devices are still limited to the observation of DC acoustoelectric currents, and have yet to achieve real-time electronic signal processing capabilities. In this paper, we have designed a hybrid acoustoelectric platform composed of two-dimensional semiconductor FET and SAW device. The platform is capable of processing DC signals, exhibiting ambipolar transport behavior. The sub-wavelength channel length of the FET within the platform allows for the real-time observation of carrier distribution at a microscopic scale in conjunction with the SAW potential, and facilitating the reproduction and intensity regulation of AC signals. By adjusting the relative phase and intensity ratio of two counter-propagating SAWs, the platform also enables the addition and subtraction of AC signals.

physics.app-ph↗

arXiv2025

Boosting the Self-driven Properties of 2D Photodetectors through Synergistic Asymmetrical Effects

Self-driven photodetectors (SDPDs) transform photon energy into electrical energy without external voltage, which makes them highly advantageous for applications such as low-power communication and imaging systems. Two-dimensional materials (2DMs) provide ideal platforms for SDPDs thanks to their band structures covering ultraviolet to infrared spectrum, strong light absorption efficiencies, and high carrier mobilities. However, the lack of stable doping methods and the complicated 2DMs multilayer stacking techniques pose tremendous difficulties for 2DMs to adopt the same device structures (i.e. PN junctions) as bulk materials, and the resultant self-driven performance remains at a low level. This work reveals how different asymmetrical effects can be combined to synergistically boost self-driven properties based on typical 2D metal-semiconductor-metal (MSM) photodetectors. Using WSe2 as an exemplary 2D material to build MSM photodetectors, the synergistic effect of asymmetrical contact electrodes and asymmetrical contact geometries is theoretically and experimentally demonstrated. The open-circuit voltage (Voc) of the SDPD reaches 0.58V, with a zero-bias responsivity of 5.77 A/W and an on/off ratio of 1.73*10^5. Additionally, our devices demonstrate potential for visible light communication (VLC) in underwater environments. Our results offer a promising and efficient strategy for building SDPDs based on various 2DMs and pave the way toward low-power optoelectronic applications.

physics.optics↗

arXiv2024

Video Creation by Demonstration

We explore a novel video creation experience, namely Video Creation by Demonstration. Given a demonstration video and a context image from a different scene, we generate a physically plausible video that continues naturally from the context image and carries out the action concepts from the demonstration. To enable this capability, we present $\delta$-Diffusion, a self-supervised training approach that learns from unlabeled videos by conditional future frame prediction. Unlike most existing video generation controls that are based on explicit signals, we adopts the form of implicit latent control for maximal flexibility and expressiveness required by general videos. By leveraging a video foundation model with an appearance bottleneck design on top, we extract action latents from demonstration videos for conditioning the generation process with minimal appearance leakage. Empirically, $\delta$-Diffusion outperforms related baselines in terms of both human preference and large-scale machine evaluations, and demonstrates potentials towards interactive world simulation. Sampled video generation results are available at https://delta-diffusion.github.io/.

cs.CV↗

arXiv2024

Learning 3D Perception from Others' Predictions

Accurate 3D object detection in real-world environments requires a huge amount of annotated data with high quality. Acquiring such data is tedious and expensive, and often needs repeated effort when a new sensor is adopted or when the detector is deployed in a new environment. We investigate a new scenario to construct 3D object detectors: learning from the predictions of a nearby unit that is equipped with an accurate detector. For example, when a self-driving car enters a new area, it may learn from other traffic participants whose detectors have been optimized for that area. This setting is label-efficient, sensor-agnostic, and communication-efficient: nearby units only need to share the predictions with the ego agent (e.g., car). Naively using the received predictions as ground-truths to train the detector for the ego car, however, leads to inferior performance. We systematically study the problem and identify viewpoint mismatches and mislocalization (due to synchronization and GPS errors) as the main causes, which unavoidably result in false positives, false negatives, and inaccurate pseudo labels. We propose a distance-based curriculum, first learning from closer units with similar viewpoints and subsequently improving the quality of other units' predictions via self-training. We further demonstrate that an effective pseudo label refinement module can be trained with a handful of annotated data, largely reducing the data quantity necessary to train an object detector. We validate our approach on the recently released real-world collaborative driving dataset, using reference cars' predictions as pseudo labels for the ego car. Extensive experiments including several scenarios (e.g., different sensors, detectors, and domains) demonstrate the effectiveness of our approach toward label-efficient learning of 3D perception from other units' predictions.

cs.CV↗

arXiv2024

MOD-UV: Learning Mobile Object Detectors from Unlabeled Videos

Embodied agents must detect and localize objects of interest, e.g. traffic participants for self-driving cars. Supervision in the form of bounding boxes for this task is extremely expensive. As such, prior work has looked at unsupervised instance detection and segmentation, but in the absence of annotated boxes, it is unclear how pixels must be grouped into objects and which objects are of interest. This results in over-/under-segmentation and irrelevant objects. Inspired by human visual system and practical applications, we posit that the key missing cue for unsupervised detection is motion: objects of interest are typically mobile objects that frequently move and their motions can specify separate instances. In this paper, we propose MOD-UV, a Mobile Object Detector learned from Unlabeled Videos only. We begin with instance pseudo-labels derived from motion segmentation, but introduce a novel training paradigm to progressively discover small objects and static-but-mobile objects that are missed by motion segmentation. As a result, though only learned from unlabeled videos, MOD-UV can detect and segment mobile objects from a single static image. Empirically, we achieve state-of-the-art performance in unsupervised mobile object detection on Waymo Open, nuScenes, and KITTI Datasets without using any external data or supervised models. Code is available at https://github.com/YihongSun/MOD-UV.

cs.CV↗

arXiv2023

Dynamo-Depth: Fixing Unsupervised Depth Estimation for Dynamical Scenes

Unsupervised monocular depth estimation techniques have demonstrated encouraging results but typically assume that the scene is static. These techniques suffer when trained on dynamical scenes, where apparent object motion can equally be explained by hypothesizing the object's independent motion, or by altering its depth. This ambiguity causes depth estimators to predict erroneous depth for moving objects. To resolve this issue, we introduce Dynamo-Depth, an unifying approach that disambiguates dynamical motion by jointly learning monocular depth, 3D independent flow field, and motion segmentation from unlabeled monocular videos. Specifically, we offer our key insight that a good initial estimation of motion segmentation is sufficient for jointly learning depth and independent motion despite the fundamental underlying ambiguity. Our proposed method achieves state-of-the-art performance on monocular depth estimation on Waymo Open and nuScenes Dataset with significant improvement in the depth of moving objects. Code and additional results are available at https://dynamo-depth.github.io.

cs.CV↗

arXiv2022

Amodal Segmentation through Out-of-Task and Out-of-Distribution Generalization with a Bayesian Model

Amodal completion is a visual task that humans perform easily but which is difficult for computer vision algorithms. The aim is to segment those object boundaries which are occluded and hence invisible. This task is particularly challenging for deep neural networks because data is difficult to obtain and annotate. Therefore, we formulate amodal segmentation as an out-of-task and out-of-distribution generalization problem. Specifically, we replace the fully connected classifier in neural networks with a Bayesian generative model of the neural network features. The model is trained from non-occluded images using bounding box annotations and class labels only, but is applied to generalize out-of-task to object segmentation and to generalize out-of-distribution to segment occluded objects. We demonstrate how such Bayesian models can naturally generalize beyond the training task labels when they learn a prior that models the object's background context and shape. Moreover, by leveraging an outlier process, Bayesian models can further generalize out-of-distribution to segment partially occluded objects and to predict their amodal object boundaries. Our algorithm outperforms alternative methods that use the same supervision by a large margin, and even outperforms methods where annotated amodal segmentations are used during training, when the amount of occlusion is large. Code is publicly available at https://github.com/YihongSun/Bayesian-Amodal.

cs.CV↗

arXiv2021

Robust Instance Segmentation through Reasoning about Multi-Object Occlusion

Analyzing complex scenes with Deep Neural Networks is a challenging task, particularly when images contain multiple objects that partially occlude each other. Existing approaches to image analysis mostly process objects independently and do not take into account the relative occlusion of nearby objects. In this paper, we propose a deep network for multi-object instance segmentation that is robust to occlusion and can be trained from bounding box supervision only. Our work builds on Compositional Networks, which learn a generative model of neural feature activations to locate occluders and to classify objects based on their non-occluded parts. We extend their generative model to include multiple objects and introduce a framework for efficient inference in challenging occlusion scenarios. In particular, we obtain feed-forward predictions of the object classes and their instance and occluder segmentations. We introduce an Occlusion Reasoning Module (ORM) that locates erroneous segmentations and estimates the occlusion order to correct them. The improved segmentation masks are, in turn, integrated into the network in a top-down manner to improve the image classification. Our experiments on the KITTI INStance dataset (KINS) and a synthetic occlusion dataset demonstrate the effectiveness and robustness of our model at multi-object instance segmentation under occlusion. Code is publically available at https://github.com/XD7479/Multi-Object-Occlusion.

cs.CV↗

arXiv2020

Compositional Convolutional Neural Networks: A Robust and Interpretable Model for Object Recognition under Occlusion

Computer vision systems in real-world applications need to be robust to partial occlusion while also being explainable. In this work, we show that black-box deep convolutional neural networks (DCNNs) have only limited robustness to partial occlusion. We overcome these limitations by unifying DCNNs with part-based models into Compositional Convolutional Neural Networks (CompositionalNets) - an interpretable deep architecture with innate robustness to partial occlusion. Specifically, we propose to replace the fully connected classification head of DCNNs with a differentiable compositional model that can be trained end-to-end. The structure of the compositional model enables CompositionalNets to decompose images into objects and context, as well as to further decompose object representations in terms of individual parts and the objects' pose. The generative nature of our compositional model enables it to localize occluders and to recognize objects based on their non-occluded parts. We conduct extensive experiments in terms of image classification and object detection on images of artificially occluded objects from the PASCAL3D+ and ImageNet dataset, and real images of partially occluded vehicles from the MS-COCO dataset. Our experiments show that CompositionalNets made from several popular DCNN backbones (VGG-16, ResNet50, ResNext) improve by a large margin over their non-compositional counterparts at classifying and detecting partially occluded objects. Furthermore, they can localize occluders accurately despite being trained with class-level supervision only. Finally, we demonstrate that CompositionalNets provide human interpretable predictions as their individual components can be understood as detecting parts and estimating an objects' viewpoint.

cs.CV↗

arXiv2020

Robust Object Detection under Occlusion with Context-Aware CompositionalNets

Detecting partially occluded objects is a difficult task. Our experimental results show that deep learning approaches, such as Faster R-CNN, are not robust at object detection under occlusion. Compositional convolutional neural networks (CompositionalNets) have been shown to be robust at classifying occluded objects by explicitly representing the object as a composition of parts. In this work, we propose to overcome two limitations of CompositionalNets which will enable them to detect partially occluded objects: 1) CompositionalNets, as well as other DCNN architectures, do not explicitly separate the representation of the context from the object itself. Under strong object occlusion, the influence of the context is amplified which can have severe negative effects for detection at test time. In order to overcome this, we propose to segment the context during training via bounding box annotations. We then use the segmentation to learn a context-aware CompositionalNet that disentangles the representation of the context and the object. 2) We extend the part-based voting scheme in CompositionalNets to vote for the corners of the object's bounding box, which enables the model to reliably estimate bounding boxes for partially occluded objects. Our extensive experiments show that our proposed model can detect objects robustly, increasing the detection performance of strongly occluded vehicles from PASCAL3D+ and MS-COCO by 41% and 35% respectively in absolute performance relative to Faster R-CNN.

cs.CV↗