论文检索

Yiqian Chang, Haoran Xu, Qinghong Ye, Jianing Li, Xuan Wang, Wei Zhang, Peixi Peng

High spatio‑temporal resolution novel‑view scene rendering is crucial for applications such as sports analysis and scientific experiments. However, existing Dynamic Scene Rendering (DSR) approaches typically rely on conventional RGB cameras with limited frame rates, making it difficult to achieve high spatio‑temporal resolution. In this paper, we present BulletTime4D, a high spatio‑temporal resolution DSR framework, which is the first trial to integrate a spike camera with binocular RGB cameras for dynamic scene reconstruction. Specifically, we first develop a hybrid camera prototype and build a real‑world dynamic scene reconstruction dataset. Then, BulletTime4D presents a multi‑timescale deformation representation by combining low‑frequency spatio‑temporal features with high‑frequency inter‑frame motion features. Finally, a rendering network is designed capable of projecting 4D Gaussians into the spike domain for spike rendering, and a cross‑domain supervision strategy is proposed to achieve high‑frame‑rate texture and color rendering. The results show that BulletTime4D outperforms state‑of‑the‑art methods on both simulated and real‑world datasets. In addition, BulletTime4D can synthesize 300 FPS novel‑view renderings using stereo RGB cameras at 30 FPS and a single spike camera.

Luyao Chang, Leiting Chen, Chen Yang, Chuan Zhou

Weakly supervised semantic segmentation (WSSS) suffers from an inherent mismatch between coarse image-level annotations and dense pixel-level predictions. To bridge this gap, existing methods primarily focus on generating refined class activation maps (CAM) as pseudo-labels. However, we argue that this focus is insufficient as it overlooks a critical component: the segmentation decoder. The decoder is typically trained through superficial alignment of predictions with pseudo-labels in the logit space. Given the noisy nature of such labels, this naive supervision leads to error accumulation and limits performance. To address this, we propose an Uncertainty-Guided Reliable Learning (UGRL) framework that exerts dual control to reshape the learning process, achieving robust supervision that escapes the CAM shadow. The cornerstone of UGRL is a prototype-driven uncertainty modeling module that estimates the reliability of class-wise supervision. The modeled uncertainty enables two synergistic control mechanisms. First, it adaptively modulates classification and segmentation losses, encouraging the model to learn from more trustworthy signals. Second, it guides the structuring of the decoder’s feature space. Rather than relying solely on superficial alignment, UGRL enforces deeper representation alignment by applying contrastive learning on reliable pixels. This enables rich semantic transfer to fine-grained segmentation details. Extensive experiments on PASCAL VOC and MS COCO demonstrate that our method surpasses other state-of-the-art WSSS methods.

Hua Chang, Xin Xu, Wei Liu, Wei Wang, Xin Yuan, Kui Jiang

Chinese opera is celebrated for preserving classical art. However, early filming equipment limitations have degraded videos of last-century performances by renowned artists (e.g., low frame rates and resolution), hindering archival efforts. Although space-time video super-resolution (STVSR) has advanced significantly, applying it directly to opera videos remains challenging. The scarcity of datasets impedes the recovery of high-frequency details, and existing STVSR methods lack global modeling capabilities—compromising visual quality when handling opera’s characteristic large motions. To address these challenges, we pioneer a large-scale Chinese Opera Video Clip (COVC) dataset and propose the Mamba-based multiscale fusion network for space-time Opera Video Super-Resolution (MambaOVSR). Specifically, MambaOVSR involves three novel components: the Global Fusion Module (GFM) for motion modeling through a multiscale alternating scanning mechanism, and the Multiscale Synergistic Mamba Module (MSMM) for alignment across different sequence lengths. Additionally, our MambaVR block resolves feature artifacts and positional information loss during alignment. Experimental results on the COVC dataset show that MambaOVSR significantly outperforms the SOTA STVSR method by an average of 1.86 dB in terms of PSNR.

Hanzhi Chang, Ruijie Zhu, Wenjie Chang, Mulin Yu, Yanzhe Liang, Jiahao Lu, Zhuoyuan Li, Tianzhu Zhang

Surface reconstruction has been widely studied in computer vision and graphics. However, existing surface reconstruction works struggle to recover accurate scene geometry when the input views are extremely sparse. To address this issue, we propose MeshSplat, a generalizable sparse-view surface reconstruction framework via Gaussian Splatting. Our key idea is to leverage 2DGS as a bridge, which connects novel view synthesis to learned geometric priors and then transfers these priors to achieve surface reconstruction. Specifically, we incorporate a feed-forward network to predict per-view pixel-aligned 2DGS, which enables the network to synthesize novel view images and thus eliminates the need for direct 3D ground-truth supervision. To improve the accuracy of 2DGS position and orientation prediction, we propose a Weighted Chamfer Distance Loss to regularize the depth maps, especially in overlapping areas of input views, and also a normal prediction network to align the orientation of 2DGS with normal vectors predicted by a monocular normal estimator. Extensive experiments validate the effectiveness of our proposed improvement, demonstrating that our method achieves state-of-the-art performance in generalizable sparse-view mesh reconstruction tasks.

Gyusam Chang, Tuan-Anh Vu, Vivek Alumootil, Harris Song, Deanna Pham, Sangpil Kim, M. Khalid Jawed

While 3D Gaussian Splatting (3DGS) has rapidly advanced, its application in agriculture remains underexplored. Agricultural scenes pose unique challenges for 3D reconstruction methods, notably uneven illumination, occlusions, and limited perspectives. To address these limitations, we introduce NTRPlant, a novel multimodal dataset encompassing Near-Infrared (NIR), RGB imagery, textual metadata, Depth, and LiDAR collected under varied indoor and outdoor lighting conditions. By integrating NIR data, our approach enhances robustness and extracts crucial botanical insights beyond visible spectra. Additionally, we leverage text-based metadata derived from vegetation indices, such as NDVI, NDWI, and chlorophyll index, significantly enriching the contextual understanding of complex agricultural environments. To fully exploit these modalities, we propose NIRSplat, an effective multimodal Gaussian splatting architecture employing a cross-attention mechanism combined with 3D point-based positional encoding, providing robust geometric priors. Comprehensive experiments demonstrate that NIRSplat outperforms existing state-of-the-art methods, including 3DGS and InstantSplat, highlighting its effectiveness in challenging agricultural scenarios.

Boyu Chang, Qi Wang, Xi Guo, Zhixiong Nan, Yazhou Yao, Tianfei Zhou

Visual abductive reasoning (VAR) is a challenging task that requires AI systems to infer the most likely explanation for incomplete visual observations. While recent MLLMs develop strong general-purpose multimodal reasoning capabilities, they remain fall short in abductive inference, as compared to human beings. To bridge this gap, we draw inspiration from the interplay between verbal and pictorial abduction in human cognition, and propose to strengthen abduction of MLLMs by mimicking such dual-mode behavior. Concretely, we introduce AbductiveMLLM comprising of two synergistic components: REASONER and IMAGINER. The REASONER operates in the verbal domain. It first explores a broad space of possible explanations using a blind LLM and then prunes visually incongruent hypotheses based on cross-modal causal alignment. The remaining hypotheses are introduced into the MLLM as targeted priors, steering its reasoning toward causally coherent explanations. The IMAGINER, on the other hand, further guides MLLMs by emulating human-like pictorial thinking. It conditions a text-to-image diffusion model on both the input video and the REASONER’s output embeddings to “imagine” plausible visual scenes that correspond to verbal explanation, thereby enriching MLLMs' contextual grounding. The two components are trained jointly in an end-to-end manner. Experiments on standard VAR benchmarks show that AbductiveMLLM achieves state-of-the-art performance, consistently outperforming traditional solutions and advanced MLLMs.

Bowen Chai, Zheng Chen, Libo Zhu, Wenbo Li, Yong Guo, Yulun Zhang

Diffusion models have shown superior performance in real-world video super-resolution (VSR). However, the slow processing speeds and heavy resource consumption of diffusion models hinder their practical application and deployment. Quantization offers a potential solution for compressing the VSR model. Nevertheless, quantizing VSR models is challenging due to their temporal characteristics and high fidelity requirements. To address these issues, we propose QuantVSR, a low-bit quantization model for real-world VSR. We propose a spatio-temporal complexity aware (STCA) mechanism, where we first utilize the calibration dataset to measure both spatial and temporal complexities for each layer. Based on these statistics, we allocate layer-specific ranks to the low-rank full-precision (FP) auxiliary branch. Subsequently, we jointly refine the FP and low-bit branches to achieve simultaneous optimization. In addition, we propose a learnable bias alignment (LBA) module to reduce the biased quantization errors. Extensive experiments on synthetic and real-world datasets demonstrate that our method obtains comparable performance with the FP model and significantly outperforms recent leading low-bit quantization methods.

Zijian Cao, Dayou Zhang, Zeyuan Liu, Zhicheng Liang, Fangxin Wang

Point-based geometric representations such as point clouds and Gaussian Splatting are fundamental for 3D understanding. However, the inherent irregularity and high-dimensional nature of point structures present significant challenges for direct 3D learning approaches, which often struggle with scalability and achieve suboptimal performance due to sparse data distributions. In contrast, 2D learning paradigms benefit from well-established architectures with superior optimization stability and efficiency. To bridge this gap, we propose Maniflat3D, a unified framework that systematically transforms volumetric point-based geometries into structured 2D representations through a two-stage process: a multilayer Ball-Pivoting reconstruction with adaptive density control, followed by Scalable Locally Injective Mapping (SLIM) to produce distortion-minimized, bijective UV parameterizations. Our approach explicitly encodes both geometric and attribute information into the flattened domain, enabling conventional 2D neural networks to effectively learn from complex 3D structures such as Gaussian Splatting. Experiments on the ShapeSplat dataset demonstrate that Maniflat3D achieves comparable performance while reducing parameter count by 90% compared to native 3D baselines, and simultaneously attains 21× compression ratio through neural encoding. These results establish a new paradigm for efficient geometric understanding, demonstrating successful transfer of planar learning advantages to challenging 3D manifold problems through dimensional reduction.

Yushe Cao, Dianxi Shi, Xing Fu, Xuechao Zou, Haikuo Peng, Xueqi Li, Chun Yu, Junliang Xing

While significant progress has been achieved in multimodal facial generation using semantic masks and textual descriptions, conventional feature fusion approaches often fail to enable effective cross-modal interactions, thereby leading to suboptimal generation outcomes. To address this challenge, we introduce MDiTFace—a customized diffusion transformer framework that employs a unified tokenization strategy to process semantic mask and text inputs, eliminating discrepancies between heterogeneous modality representations. The framework facilitates comprehensive multimodal feature interaction through stacked, newly designed multivariate transformer blocks that process all conditions synchronously. Additionally, we design a novel decoupled attention mechanism by dissociating implicit dependencies between mask tokens and temporal embeddings. This mechanism segregates internal computations into dynamic and static pathways, enabling caching and reuse of features computed in static pathways after initial calculation, thereby reducing additional computational overhead introduced by mask condition by over 94% while maintaining performance. Extensive experiments demonstrate that MDiTFace significantly outperforms other competing methods in terms of both facial fidelity and conditional consistency.

Yuqin Cao, Yixuan Gao, Wei Sun, Xiaohong Liu, Yulun Zhang, Xiongkuo Min

Face videos accompanied by audio have become integral to our daily lives, while they often suffer from complex degradations. Most face video restoration methods neglect the intrinsic correlations between visual and audio features, particularly in the mouth region. Several audio-aided face video restoration methods have been proposed, but they only focus on compression artifact removal. In this paper, we propose a General Audio-assisted face Video restoration Network (GAVN) to address various types of streaming video distortions via identity and temporal complementary learning. Specifically, GAVN first captures inter-frame temporal features in the low-resolution space to restore frames coarsely and save computational cost. Then, GAVN extracts intra-frame identity features in the high-resolution space with the assistance of audio signals and face landmarks to restore more facial details. Finally, the reconstruction module integrates temporal features and identity features to generate high-quality face videos. Experimental results demonstrate that GAVN outperforms the existing state-of-the-art methods on face video compression artifact removal, deblurring, and super-resolution.

Yue Cao, Zhuo Zhang, Shuai Xiao, Jialin Li, Guipeng Lan, Jiabao Wen, Jiachen Yang

Multi-view automatic translational correction (ATC) in coronary angiography (CAG) is critical for intraoperative automatic diagnosis, in which deep learning playing a key role. However, heartbeat-induced soft matching errors and costly annotations make it difficult to build high-quality, large-scale datasets for calibration algorithm training. The training of clinical models is difficult to fulfill, as existing datasets differ significantly from real CAG in both style and structure. To address this challenge, we propose a novel high-quality data synthesis method for annotation-free ATC. We fully automated the construction of a labeled, high-fidelity dataset for training matching models. An evolutionary algorithm is introduced for global optimization of translation estimation, mitigating epipolar constraint violations caused by vascular deformation and enabling reliable correction across large viewpoint differences. Furthermore, a theoretical analysis is presented, demonstrating that error propagation between adjacent views is more accurate than direct estimation across distant views. Our experiments on clinical datasets demonstrate that our method not only significantly outperforms weakly supervised learning approaches, but also performs comparably to fully supervised methods. Moreover, it exhibits remarkable multicenter generalizability.

Yue Cao, Sizhao Li, Liguo Zhang

The inherent differences between spike cameras and traditional frame-based cameras lead to more complex and diverse noise characteristics, particularly under extremely low-light conditions. Existing noise modeling approaches for spike camera predominantly rely on inter-spike intervals (ISI) for noise quantification, which often results in inaccurate noise characterization. Moreover, current datasets for spike camera image reconstruction tasks are either synthetic or lack corresponding high-quality reference images, severely limiting rigorous evaluation of noise modeling methods. To address this limitation, we propose a multimodal noise modeling framework for spike camera that integrates insights from traditional frame-based imaging into spike imaging. Specifically, we introduce a time-interval-based quantification method inspired by the exposure-time concept used in traditional frame-based cameras, enabling accurate noise characterization for spike camera. Furthermore, we present the Spike-DSLR Multimodal Dataset (SDMD), the first real-world dataset capturing aligned multimodal data pairs from spike cameras and Digital Single-Lens Reflex (DSLR) cameras, explicitly designed for evaluating spike camera noise models. Experimental results on SDMD demonstrate that our noise modeling approach significantly enhances spike camera image reconstruction quality under low-light conditions, achieving more than 1.6 dB improvement in PSNR compared to existing state-of-the-art methods. This validates both the necessity and effectiveness of adopting a multimodal perspective in spike camera noise modeling.

Yifei Cao, Yu Liu, Guolong Wang, Zhu Liu, Kai Wang, Xianjie Zhang, Jizhe Yu, Xun Tu

Egocentric visual query localization is vital for embodied AI and VR/AR, yet remains challenging due to camera motion, viewpoint changes, and appearance variations. We present EAGLE, a novel framework that leverages episodic appearance- and geometry-aware memory to achieve unified 2D-3D visual query localization in egocentric vision. Inspired by avian memory consolidation, EAGLE synergistically integrates segmentation guided by an appearance-aware meta-learning memory (AMM), with tracking driven by a geometry-aware localization memory (GLM). This memory consolidation mechanism, through structured appearance and geometry memory banks, stores high-confidence retrieval samples, effectively supporting both long- and short-term modeling of target appearance variations. This enables precise contour delineation with robust spatial discrimination, leading to significantly improved retrieval accuracy. Furthermore, by integrating the VQL-2D output with a visual geometry grounded Transformer (VGGT), we achieve a efficient unification of 2D and 3D tasks, enabling rapid and accurate back-projection into 3D space. Our method achieves state-of-the-art performance on the Ego4D-VQ benchmark.

Qinglong Cao, Xirui Li, Ding Wang, Chao Ma, Yuntian Chen, Xiaokang Yang

Video diffusion models have achieved impressive results in natural scene generation, yet they struggle to generalize to scientific phenomena such as fluid simulations and meteorological processes, where underlying dynamics are governed by scientific laws. These tasks pose unique challenges, including severe domain gaps, limited training data, and the lack of descriptive language annotations. To handle this dilemma, we extracted the latent scientific phenomena knowledge and further proposed a fresh framework that teaches video diffusion models to generate scientific phenomena from a single initial frame. Particularly, static knowledge is extracted via pre-trained masked autoencoders, while dynamic knowledge is derived from pre-trained optical flow prediction. Subsequently, based on the aligned spatial relations between the CLIP vision and language encoders, the visual embeddings of scientific phenomena, guided by latent scientific phenomena knowledge, are projected to generate the pseudo-language prompt embeddings in both spatial and frequency domains. By incorporating these prompts and fine-tuning the video diffusion model, we enable the generation of videos that better adhere to scientific laws. Extensive experiments on both computational fluid dynamics simulations and real-world typhoon observations demonstrate the effectiveness of our approach, achieving superior fidelity and consistency across diverse scientific scenarios.

Meng Cao, Pengfei Hu, Yingyao Wang, Jihao Gu, Haoran Tang, Haoze Zhao, Chen Wang, Jiahua Dong, Wangbo Yu, Ge Zhang 等

Recent advancements in Large Video Language Models (LVLMs) have highlighted their potential for multi-modal understanding, yet evaluating their factual grounding in videos remains a critical unsolved challenge. To address this gap, we introduce Video SimpleQA, the first comprehensive benchmark tailored for factuality evaluation in video contexts. Our work differs from existing video benchmarks through the following key features: 1) Knowledge required: demanding integration of external knowledge beyond the video’s explicit narrative; 2) Multi-hop fact-seeking question: Each question involves multiple explicit facts and requires strict factual grounding without hypothetical or subjective inferences. We include per-hop single-fact-based sub-QAs alongside final QAs to enable fine-grained, step-by-step evaluation; 3) Short-form definitive answer: Answers are crafted as unambiguous and definitively correct in a short format with minimal scoring variance; 4) Temporal grounded required: Requiring answers to rely on one or more temporal segments in videos, rather than single frames. We extensively evaluate 33 state-of-the-art LVLMs and summarize key findings as follows: 1) Current LVLMs exhibit notable deficiencies in factual adherence, with the best-performing model o3 merely achieving an F-score of 66.3%; 2) Most LVLMs are overconfident in what they generate, with self-stated confidence exceeding actual accuracy; 3) Retrieval-Augmented Generation demonstrates consistent improvements at the cost of additional inference time overhead; 4) Multi-hop QA demonstrates substantially degraded performance compared to single-hop sub-QAs, with first-hop object/event recognition emerging as the primary bottleneck. We position Video SimpleQA as the cornerstone benchmark for video factuality assessment, aiming to steer LVLM development toward verifiable grounding in real-world contexts.

Linhan Cao, Wei Sun, Weixia Zhang, Xiangyang Zhu, Jun Jia, Kaiwei Zhang, Dandan Zhu, Guangtao Zhai, Xiongkuo Min

Video quality assessment (VQA) aims to objectively quantify perceptual quality degradation in alignment with human visual perception. Despite recent advances, existing VQA models still suffer from two critical limitations: poor generalization to out-of-distribution (OOD) videos and limited explainability, which restrict their applicability in real-world scenarios. To address these challenges, we propose VQAThinker, a reasoning-based VQA framework that leverages large multimodal models (LMMs) with reinforcement learning to jointly model video quality understanding and scoring, emulating human perceptual decision-making. Specifically, we adopt group relative policy optimization (GRPO), a rule-guided reinforcement learning algorithm that enables reasoning over video quality under score-level supervision, and introduce three VQA-specific rewards: (1) a bell-shaped regression reward that increases rapidly as the prediction error decreases and becomes progressively less sensitive near the ground truth; (2) a pairwise ranking reward that guides the model to correctly determine the relative quality between video pairs; and (3) a temporal consistency reward that encourages the model to prefer temporally coherent videos over their perturbed counterparts. Extensive experiments demonstrate that VQAThinker achieves state-of-the-art performance on both in-domain and OOD VQA benchmarks, showing strong generalization for video quality scoring. Furthermore, evaluations on video quality understanding tasks validate its superiority in distortion attribution and quality description compared to existing explainable VQA models and LMMs. These findings demonstrate that reinforcement learning offers an effective pathway toward building generalizable and explainable VQA models solely with score-level supervision.

Ke Cao, Jing Wang, Ao Ma, Jiasong Feng, Xuanhua He, Run Ling, Haowei Liu, Jian Lu, Wei Feng, Haozhe Wang 等

The Diffusion Transformer plays a pivotal role in advancing text-to-image and text-to-video generation, owing primarily to its inherent scalability. However, existing controlled diffusion transformer methods incur significant parameter and computational overheads and suffer from inefficient resource allocation due to their failure to account for the varying relevance of control information across different transformer layers. To address this, we propose the Relevance-Guided Efficient Controllable Generation framework, RelaCtrl, enabling efficient and resource-optimized integration of control signals into the Diffusion Transformer. First, we evaluate the relevance of each layer in the Diffusion Transformer to the control information by assessing the ControlNet Relevance Score, which measures the impact of skipping each control layer on both the quality of generation and the control effectiveness during inference. Based on the strength of the relevance, we then tailor the positioning, parameter scale, and modeling capacity of the control layers to reduce unnecessary parameters and redundant computations. Additionally, to further improve efficiency, we replace the self-attention and FFN in the commonly used copy block with the carefully designed Two-Dimensional Shuffle Mixer (TDSM), enabling efficient implementation of both the token mixer and channel mixer. Both qualitative and quantitative experimental results demonstrate that our approach achieves superior performance with only 15% of the parameters and computational complexity compared to PixArt-delta.

Jingchao Cao, Guo An, Feng Gao, Ke Gu, Yutao Liu

Despite the rapid progress of multimodal large language models (MLLMs), their capacity for low-level visual perception in underwater environments remains underexplored. To address this gap, we present UQ-Bench, the first systematically designed benchmark for evaluating the ability of MLLMs to perceive and assess underwater image quality at the low-level visual attribute level. UQ-Bench comprises three components: (1) UW-Perception, a dataset of 3,000 underwater images paired with targeted questions on key degradations such as color cast, blur, contrast, and exposure, covering both global and local perceptual dimensions; (2) UW-Describe, a dataset of 500 images with expert-annotated gold-standard descriptions for assessing the accuracy of model-generated text; and (3) UW-Eval, an evaluation protocol employing human mean opinion scores (MOS) for quantitative quality assessment. To ensure rigorous and reproducible benchmarking, we propose a GPT-assisted evaluation framework that aligns model outputs with expert references and enables fine-grained analysis of distortion perception. Experimental results demonstrate that while MLLMs exhibit preliminary competence in underwater low-level visual tasks, they still fall short in capturing subtle degradations and achieving human-level consistency, highlighting the need for further advances in foundation models for marine vision.

Jing Cao, Kui Jiang, Shenyi Li, Xiaocheng Feng, Yong Huang

Self-supervised depth estimation has gained significant attention in autonomous driving and robotics. However, existing methods exhibit substantial performance degradation under adverse weather conditions such as rain and fog, where reduced visibility critically impairs depth prediction. To address this issue, we propose a novel self-evolution contrastive learning framework called SEC-Depth for self-supervised robust depth estimation tasks. Our approach leverages intermediate parameters generated during training to construct temporally evolving latency models. Using these, we design a self-evolution contrastive scheme to mitigate performance loss under challenging conditions. Concretely, we first design a dynamic update strategy of latency models for the depth estimation task to capture optimization states across training stages. To effectively leverage latency models, we introduce a self-evolution contrastive Loss (SECL) that treats outputs from historical latency models as negative samples. This mechanism adaptively adjusts learning objectives while implicitly sensing weather degradation severity, reducing the needs for manual intervention. Experiments show that our method integrates seamlessly into diverse baseline models and significantly enhances robustness in zero-shot evaluations.

Jiajun Cao, Qizhe Zhang, Peidong Jia, Xuhui Zhao, Bo Lan, Xiaoan Zhang, Lizhuo, Xiaobao Wei, Sixiang Chen, Liyun Li 等

Vision-Language-Action (VLA) models have demonstrated significant potential in complex scene understanding and action reasoning, leading to their increasing adoption in end-to-end autonomous driving systems. However, the long visual tokens of VLA models greatly increase computational costs. Current visual token pruning methods in Vision-Language Models (VLM) rely on either visual token similarity or visual-text attention, but both have shown poor performance in autonomous driving scenarios. Given that human drivers concentrate on relevant foreground areas while driving, we assert that retaining visual tokens containing this foreground information is essential for effective decision-making. Inspired by this, we propose FastDriveVLA, a novel reconstruction-based vision token pruning framework designed specifically for autonomous driving. FastDriveVLA includes a plug-and-play visual token pruner called ReconPruner, which prioritizes foreground information through MAE-style pixel reconstruction. A novel adversarial foreground-background reconstruction strategy is designed to train ReconPruner for the visual encoder of VLA models. Once trained, ReconPruner can be seamlessly applied to different VLA models with the same visual encoder without retraining. To train ReconPruner, we also introduce a large-scale dataset called nuScenes-FG, consisting of 241K image-mask pairs with annotated foreground regions. Our approach achieves state-of-the-art results on the nuScenes open-loop planning benchmark across different pruning ratios.