Seaweed-7B: Cost-Effective Training of Video Generation Foundation Model

This technical report presents a cost-efficient strategy for training a video generation foundation model. We present a mid-sized research model with approximately 7 billion parameters (7B) called Seaweed-7B trained from scratch using 665,000 H100 GPU hours. Despite being trained with moderate computational resources, Seaweed-7B demonstrates highly competitive performance compared to contemporary video generation models of much larger size. Design choices are especially crucial in a resource-constrained setting. This technical report highlights the key design decisions that enhance the performance of the medium-sized diffusion model. Empirically, we make two observations: (1) Seaweed-7B achieves performance comparable to, or even surpasses, larger models trained on substantially greater GPU resources, and (2) our model, which exhibits strong generalization ability, can be effectively adapted across a wide range of downstream applications either by lightweight fine-tuning or continue training. See the project page at https://seaweed.video/

Digital Beamforming Enhanced Radar Odometry

Radar has become an essential sensor for autonomous navigation, especially in challenging environments where camera and LiDAR sensors fail. 4D single-chip millimeter-wave radar systems, in particular, have drawn increasing attention thanks to their ability to provide spatial and Doppler information with low hardware cost and power consumption. However, most single-chip radar systems using traditional signal processing, such as Fast Fourier Transform, suffer from limited spatial resolution in radar detection, significantly limiting the performance of radar-based odometry and Simultaneous Localization and Mapping (SLAM) systems. In this paper, we develop a novel radar signal processing pipeline that integrates spatial domain beamforming techniques, and extend it to 3D Direction of Arrival estimation. Experiments using public datasets are conducted to evaluate and compare the performance of our proposed signal processing pipeline against traditional methodologies. These tests specifically focus on assessing structural precision across diverse scenes and measuring odometry accuracy in different radar odometry systems. This research demonstrates the feasibility of achieving more accurate radar odometry by simply replacing the standard FFT-based processing with the proposed pipeline. The codes are available at GitHub*.

AQUA-SLAM: Tightly-Coupled Underwater Acoustic-Visual-Inertial SLAM with Sensor Calibration

Underwater environments pose significant challenges for visual Simultaneous Localization and Mapping (SLAM) systems due to limited visibility, inadequate illumination, and sporadic loss of structural features in images. Addressing these challenges, this paper introduces a novel, tightly-coupled Acoustic-Visual-Inertial SLAM approach, termed AQUA-SLAM, to fuse a Doppler Velocity Log (DVL), a stereo camera, and an Inertial Measurement Unit (IMU) within a graph optimization framework. Moreover, we propose an efficient sensor calibration technique, encompassing multi-sensor extrinsic calibration (among the DVL, camera and IMU) and DVL transducer misalignment calibration, with a fast linear approximation procedure for real-time online execution. The proposed methods are extensively evaluated in a tank environment with ground truth, and validated for offshore applications in the North Sea. The results demonstrate that our method surpasses current state-of-the-art underwater and visual-inertial SLAM systems in terms of localization accuracy and robustness. The proposed system will be made open-source for the community.

PanoGen++: Domain-Adapted Text-Guided Panoramic Environment Generation for Vision-and-Language Navigation

Vision-and-language navigation (VLN) tasks require agents to navigate three-dimensional environments guided by natural language instructions, offering substantial potential for diverse applications. However, the scarcity of training data impedes progress in this field. This paper introduces PanoGen++, a novel framework that addresses this limitation by generating varied and pertinent panoramic environments for VLN tasks. PanoGen++ incorporates pre-trained diffusion models with domain-specific fine-tuning, employing parameter-efficient techniques such as low-rank adaptation to minimize computational costs. We investigate two settings for environment generation: masked image inpainting and recursive image outpainting. The former maximizes novel environment creation by inpainting masked regions based on textual descriptions, while the latter facilitates agents' learning of spatial relationships within panoramas. Empirical evaluations on room-to-room (R2R), room-for-room (R4R), and cooperative vision-and-dialog navigation (CVDN) datasets reveal significant performance enhancements: a 2.44% increase in success rate on the R2R test leaderboard, a 0.63% improvement on the R4R validation unseen set, and a 0.75-meter enhancement in goal progress on the CVDN validation unseen set. PanoGen++ augments the diversity and relevance of training environments, resulting in improved generalization and efficacy in VLN tasks.

Lifelong Learning with Task-Specific Adaptation: Addressing the Stability-Plasticity Dilemma

Lifelong learning (LL) aims to continuously acquire new knowledge while retaining previously learned knowledge. A central challenge in LL is the stability-plasticity dilemma, which requires models to balance the preservation of previous knowledge (stability) with the ability to learn new tasks (plasticity). While parameter-efficient fine-tuning (PEFT) has been widely adopted in large language models, its application to lifelong learning remains underexplored. To bridge this gap, this paper proposes AdaLL, an adapter-based framework designed to address the dilemma through a simple, universal, and effective strategy. AdaLL co-trains the backbone network and adapters under regularization constraints, enabling the backbone to capture task-invariant features while allowing the adapters to specialize in task-specific information. Unlike methods that freeze the backbone network, AdaLL incrementally enhances the backbone's capabilities across tasks while minimizing interference through backbone regularization. This architectural design significantly improves both stability and plasticity, effectively eliminating the stability-plasticity dilemma. Extensive experiments demonstrate that AdaLL consistently outperforms existing methods across various configurations, including dataset choices, task sequences, and task scales.

General Scene Adaptation for Vision-and-Language Navigation

Vision-and-Language Navigation (VLN) tasks mainly evaluate agents based on one-time execution of individual instructions across multiple environments, aiming to develop agents capable of functioning in any environment in a zero-shot manner. However, real-world navigation robots often operate in persistent environments with relatively consistent physical layouts, visual observations, and language styles from instructors. Such a gap in the task setting presents an opportunity to improve VLN agents by incorporating continuous adaptation to specific environments. To better reflect these real-world conditions, we introduce GSA-VLN, a novel task requiring agents to execute navigation instructions within a specific scene and simultaneously adapt to it for improved performance over time. To evaluate the proposed task, one has to address two challenges in existing VLN datasets: the lack of OOD data, and the limited number and style diversity of instructions for each scene. Therefore, we propose a new dataset, GSA-R2R, which significantly expands the diversity and quantity of environments and instructions for the R2R dataset to evaluate agent adaptability in both ID and OOD contexts. Furthermore, we design a three-stage instruction orchestration pipeline that leverages LLMs to refine speaker-generated instructions and apply role-playing techniques to rephrase instructions into different speaking styles. This is motivated by the observation that each individual user often has consistent signatures or preferences in their instructions. We conducted extensive experiments on GSA-R2R to thoroughly evaluate our dataset and benchmark various methods. Based on our findings, we propose a novel method, GR-DUET, which incorporates memory-based navigation graphs with an environment-specific training strategy, achieving state-of-the-art results on all GSA-R2R splits.

FlashVTG: Feature Layering and Adaptive Score Handling Network for Video Temporal Grounding

Text-guided Video Temporal Grounding (VTG) aims to localize relevant segments in untrimmed videos based on textual descriptions, encompassing two subtasks: Moment Retrieval (MR) and Highlight Detection (HD). Although previous typical methods have achieved commendable results, it is still challenging to retrieve short video moments. This is primarily due to the reliance on sparse and limited decoder queries, which significantly constrain the accuracy of predictions. Furthermore, suboptimal outcomes often arise because previous methods rank predictions based on isolated predictions, neglecting the broader video context. To tackle these issues, we introduce FlashVTG, a framework featuring a Temporal Feature Layering (TFL) module and an Adaptive Score Refinement (ASR) module. The TFL module replaces the traditional decoder structure to capture nuanced video content variations across multiple temporal scales, while the ASR module improves prediction ranking by integrating context from adjacent moments and multi-temporal-scale features. Extensive experiments demonstrate that FlashVTG achieves state-of-the-art performance on four widely adopted datasets in both MR and HD. Specifically, on the QVHighlights dataset, it boosts mAP by 5.8% for MR and 3.3% for HD. For short-moment retrieval, FlashVTG increases mAP to 125% of previous SOTA performance. All these improvements are made without adding training burdens, underscoring its effectiveness. Our code is available at https://github.com/Zhuo-Cao/FlashVTG.

SuperGSeg: Open-Vocabulary 3D Segmentation with Structured Super-Gaussians

3D Gaussian Splatting has recently gained traction for its efficient training and real-time rendering. While the vanilla Gaussian Splatting representation is mainly designed for view synthesis, more recent works investigated how to extend it with scene understanding and language features. However, existing methods lack a detailed comprehension of scenes, limiting their ability to segment and interpret complex structures. To this end, We introduce SuperGSeg, a novel approach that fosters cohesive, context-aware scene representation by disentangling segmentation and language field distillation. SuperGSeg first employs neural Gaussians to learn instance and hierarchical segmentation features from multi-view images with the aid of off-the-shelf 2D masks. These features are then leveraged to create a sparse set of what we call Super-Gaussians. Super-Gaussians facilitate the distillation of 2D language features into 3D space. Through Super-Gaussians, our method enables high-dimensional language feature rendering without extreme increases in GPU memory. Extensive experiments demonstrate that SuperGSeg outperforms prior works on both open-vocabulary object localization and semantic segmentation tasks.

Unsupervised Training of a Dynamic Context-Aware Deep Denoising Framework for Low-Dose Fluoroscopic Imaging

Fluoroscopy is critical for real-time X-ray visualization in medical imaging. However, low-dose images are compromised by noise, potentially affecting diagnostic accuracy. Noise reduction is crucial for maintaining image quality, especially given such challenges as motion artifacts and the limited availability of clean data in medical imaging. To address these issues, we propose an unsupervised training framework for dynamic context-aware denoising of fluoroscopy image sequences. First, we train the multi-scale recurrent attention U-Net (MSR2AU-Net) without requiring clean data to address the initial noise. Second, we incorporate a knowledge distillation-based uncorrelated noise suppression module and a recursive filtering-based correlated noise suppression module enhanced with motion compensation to further improve motion compensation and achieve superior denoising performance. Finally, we introduce a novel approach by combining these modules with a pixel-wise dynamic object motion cross-fusion matrix, designed to adapt to motion, and an edge-preserving loss for precise detail retention. To validate the proposed method, we conducted extensive numerical experiments on medical image datasets, including 3500 fluoroscopy images from dynamic phantoms (2,400 images for training, 1,100 for testing) and 350 clinical images from a spinal surgery patient. Moreover, we demonstrated the robustness of our approach across different imaging modalities by testing it on the publicly available 2016 Low Dose CT Grand Challenge dataset, using 4,800 images for training and 1,136 for testing. The results demonstrate that the proposed approach outperforms state-of-the-art unsupervised algorithms in both visual quality and quantitative evaluation while achieving comparable performance to well-established supervised learning methods across low-dose fluoroscopy and CT imaging.

Tokens on Demand: Token Condensation as Training-free Test-time Adaptation

In this work, we introduce Token Condensation as Adaptation (TCA), a training-free approach designed to mitigate distribution shifts encountered by vision-language models (VLMs) during test-time inference. TCA bridges distribution gaps at the patch level by condensing image tokens that exhibit low attentiveness to the <cls> token. Recognizing the <cls> token may correspond to universal concepts, TCA identifies and tracks the most reliable <cls> tokens that align specifically with target classes from historical data streams. To achieve this, we propose a context token reservoir (CTR), which retains tokens with the lowest uncertainty as ``anchors" to guide the preservation of class-relevant tokens during inference. These anchors, in turn, act as token-level classifiers to correct VLM predictions and improve visual-text alignment. Utilizing anchors sampled from CTR, TCA condenses tokens through two operations: (1) pruning class-irrelevant tokens that consistently rank low across all attention heads to reach cross-head consensus on their irrelevance, and (2) merging the remaining class-ambiguous tokens into representative centers using coreset selection, maintaining linear computational complexity. As the first method to explore token efficiency in test-time adaptation, TCA consistently demonstrates superior performance across cross-dataset and out-of-distribution adaptation tasks, reducing GFLOPs by 12.2% to 48.9% while achieving accuracy improvements up to 21.4% against the strongest baseline without introducing additional parameters.