Training-free Regional Prompting for Diffusion Transformers

Diffusion models have demonstrated excellent capabilities in text-to-image generation. Their semantic understanding (i.e., prompt following) ability has also been greatly improved with large language models (e.g., T5, Llama). However, existing models cannot perfectly handle long and complex text prompts, especially when the text prompts contain various objects with numerous attributes and interrelated spatial relationships. While many regional prompting methods have been proposed for UNet-based models (SD1.5, SDXL), but there are still no implementations based on the recent Diffusion Transformer (DiT) architecture, such as SD3 and FLUX.1.In this report, we propose and implement regional prompting for FLUX.1 based on attention manipulation, which enables DiT with fined-grained compositional text-to-image generation capability in a training-free manner. Code is available at https://github.com/antonioo-c/Regional-Prompting-FLUX.

PixelGaussian: Generalizable 3D Gaussian Reconstruction from Arbitrary Views

We propose PixelGaussian, an efficient feed-forward framework for learning generalizable 3D Gaussian reconstruction from arbitrary views. Most existing methods rely on uniform pixel-wise Gaussian representations, which learn a fixed number of 3D Gaussians for each view and cannot generalize well to more input views. Differently, our PixelGaussian dynamically adapts both the Gaussian distribution and quantity based on geometric complexity, leading to more efficient representations and significant improvements in reconstruction quality. Specifically, we introduce a Cascade Gaussian Adapter to adjust Gaussian distribution according to local geometry complexity identified by a keypoint scorer. CGA leverages deformable attention in context-aware hypernetworks to guide Gaussian pruning and splitting, ensuring accurate representation in complex regions while reducing redundancy. Furthermore, we design a transformer-based Iterative Gaussian Refiner module that refines Gaussian representations through direct image-Gaussian interactions. Our PixelGaussian can effectively reduce Gaussian redundancy as input views increase. We conduct extensive experiments on the large-scale ACID and RealEstate10K datasets, where our method achieves state-of-the-art performance with good generalization to various numbers of views. Code: https://github.com/Barrybarry-Smith/PixelGaussian.

UniDrive: Towards Universal Driving Perception Across Camera Configurations

Vision-centric autonomous driving has demonstrated excellent performance with economical sensors. As the fundamental step, 3D perception aims to infer 3D information from 2D images based on 3D-2D projection. This makes driving perception models susceptible to sensor configuration (e.g., camera intrinsics and extrinsics) variations. However, generalizing across camera configurations is important for deploying autonomous driving models on different car models. In this paper, we present UniDrive, a novel framework for vision-centric autonomous driving to achieve universal perception across camera configurations. We deploy a set of unified virtual cameras and propose a ground-aware projection method to effectively transform the original images into these unified virtual views. We further propose a virtual configuration optimization method by minimizing the expected projection error between original cameras and virtual cameras. The proposed virtual camera projection can be applied to existing 3D perception methods as a plug-and-play module to mitigate the challenges posed by camera parameter variability, resulting in more adaptable and reliable driving perception models. To evaluate the effectiveness of our framework, we collect a dataset on Carla by driving the same routes while only modifying the camera configurations. Experimental results demonstrate that our method trained on one specific camera configuration can generalize to varying configurations with minor performance degradation.

V2M: Visual 2-Dimensional Mamba for Image Representation Learning

Mamba has garnered widespread attention due to its flexible design and efficient hardware performance to process 1D sequences based on the state space model (SSM). Recent studies have attempted to apply Mamba to the visual domain by flattening 2D images into patches and then regarding them as a 1D sequence. To compensate for the 2D structure information loss (e.g., local similarity) of the original image, most existing methods focus on designing different orders to sequentially process the tokens, which could only alleviate this issue to some extent. In this paper, we propose a Visual 2-Dimensional Mamba (V2M) model as a complete solution, which directly processes image tokens in the 2D space. We first generalize SSM to the 2-dimensional space which generates the next state considering two adjacent states on both dimensions (e.g., columns and rows). We then construct our V2M based on the 2-dimensional SSM formulation and incorporate Mamba to achieve hardware-efficient parallel processing. The proposed V2M effectively incorporates the 2D locality prior yet inherits the efficiency and input-dependent scalability of Mamba. Extensive experimental results on ImageNet classification and downstream visual tasks including object detection and instance segmentation on COCO and semantic segmentation on ADE20K demonstrate the effectiveness of our V2M compared with other visual backbones.

GlobalMamba: Global Image Serialization for Vision Mamba

Vision mambas have demonstrated strong performance with linear complexity to the number of vision tokens. Their efficiency results from processing image tokens sequentially. However, most existing methods employ patch-based image tokenization and then flatten them into 1D sequences for causal processing, which ignore the intrinsic 2D structural correlations of images. It is also difficult to extract global information by sequential processing of local patches. In this paper, we propose a global image serialization method to transform the image into a sequence of causal tokens, which contain global information of the 2D image. We first convert the image from the spatial domain to the frequency domain using Discrete Cosine Transform (DCT) and then arrange the pixels with corresponding frequency ranges. We further transform each set within the same frequency band back to the spatial domain to obtain a series of images before tokenization. We construct a vision mamba model, GlobalMamba, with a causal input format based on the proposed global image serialization, which can better exploit the causal relations among image sequences. Extensive experiments demonstrate the effectiveness of our GlobalMamba, including image classification on ImageNet-1K, object detection on COCO, and semantic segmentation on ADE20K.

SparseVLM: Visual Token Sparsification for Efficient Vision-Language Model Inference

In vision-language models (VLMs), visual tokens usually consume a significant amount of computational overhead, despite their sparser information density compared to text tokens. To address this, most existing methods learn a network to prune redundant visual tokens and require additional training data. Differently, we propose an efficient training-free token optimization mechanism dubbed SparseVLM without extra parameters or fine-tuning costs. Concretely, given that visual tokens complement text tokens in VLMs for linguistic reasoning, we select visual-relevant text tokens to rate the significance of vision tokens within the self-attention matrix extracted from the VLMs. Then we progressively prune irrelevant tokens. To maximize sparsity while retaining essential information, we introduce a rank-based strategy to adaptively determine the sparsification ratio for each layer, alongside a token recycling method that compresses pruned tokens into more compact representations. Experimental results show that our SparseVLM improves the efficiency of various VLMs across a range of image and video understanding tasks. In particular, LLaVA equipped with SparseVLM reduces 61% to 67% FLOPs with a compression ratio of 78% while maintaining 93% of the accuracy. Our code is available at https://github.com/Gumpest/SparseVLMs.

FactorLLM: Factorizing Knowledge via Mixture of Experts for Large Language Models

Recent research has demonstrated that Feed-Forward Networks (FFNs) in Large Language Models (LLMs) play a pivotal role in storing diverse linguistic and factual knowledge. Conventional methods frequently face challenges due to knowledge confusion stemming from their monolithic and redundant architectures, which calls for more efficient solutions with minimal computational overhead, particularly for LLMs. In this paper, we explore the FFN computation paradigm in LLMs and introduce FactorLLM, a novel approach that decomposes well-trained dense FFNs into sparse sub-networks without requiring any further modifications, while maintaining the same level of performance. Furthermore, we embed a router from the Mixture-of-Experts (MoE), combined with our devised Prior-Approximate (PA) loss term that facilitates the dynamic activation of experts and knowledge adaptation, thereby accelerating computational processes and enhancing performance using minimal training data and fine-tuning steps. FactorLLM thus enables efficient knowledge factorization and activates select groups of experts specifically tailored to designated tasks, emulating the interactive functional segmentation of the human brain. Extensive experiments across various benchmarks demonstrate the effectiveness of our proposed FactorLLM which achieves comparable performance to the source model securing up to 85% model performance while obtaining over a 30% increase in inference speed. Code: https://github.com/zhenwuweihe/FactorLLM.

LightStereo: Channel Boost Is All Your Need for Efficient 2D Cost Aggregation

We present LightStereo, a cutting-edge stereo-matching network crafted to accelerate the matching process. Departing from conventional methodologies that rely on aggregating computationally intensive 4D costs, LightStereo adopts the 3D cost volume as a lightweight alternative. While similar approaches have been explored previously, our breakthrough lies in enhancing performance through a dedicated focus on the channel dimension of the 3D cost volume, where the distribution of matching costs is encapsulated. Our exhaustive exploration has yielded plenty of strategies to amplify the capacity of the pivotal dimension, ensuring both precision and efficiency. We compare the proposed LightStereo with existing state-of-the-art methods across various benchmarks, which demonstrate its superior performance in speed, accuracy, and resource utilization. LightStereo achieves a competitive EPE metric in the SceneFlow datasets while demanding a minimum of only 22 GFLOPs, with an inference time of just 17 ms. Our comprehensive analysis reveals the effect of 2D cost aggregation for stereo matching, paving the way for real-world applications of efficient stereo systems. Code will be available at \url{https://github.com/XiandaGuo/OpenStereo}.

Instruct Large Language Models to Drive like Humans

Motion planning in complex scenarios is the core challenge in autonomous driving. Conventional methods apply predefined rules or learn from driving data to plan the future trajectory. Recent methods seek the knowledge preserved in large language models (LLMs) and apply them in the driving scenarios. Despite the promising results, it is still unclear whether the LLM learns the underlying human logic to drive. In this paper, we propose an InstructDriver method to transform LLM into a motion planner with explicit instruction tuning to align its behavior with humans. We derive driving instruction data based on human logic (e.g., do not cause collisions) and traffic rules (e.g., proceed only when green lights). We then employ an interpretable InstructChain module to further reason the final planning reflecting the instructions. Our InstructDriver allows the injection of human rules and learning from driving data, enabling both interpretability and data scalability. Different from existing methods that experimented on closed-loop or simulated settings, we adopt the real-world closed-loop motion planning nuPlan benchmark for better evaluation. InstructDriver demonstrates the effectiveness of the LLM planner in a real-world closed-loop setting. Our code is publicly available at https://github.com/bonbon-rj/InstructDriver.

$\textit{S}^3$Gaussian: Self-Supervised Street Gaussians for Autonomous Driving

Photorealistic 3D reconstruction of street scenes is a critical technique for developing real-world simulators for autonomous driving. Despite the efficacy of Neural Radiance Fields (NeRF) for driving scenes, 3D Gaussian Splatting (3DGS) emerges as a promising direction due to its faster speed and more explicit representation. However, most existing street 3DGS methods require tracked 3D vehicle bounding boxes to decompose the static and dynamic elements for effective reconstruction, limiting their applications for in-the-wild scenarios. To facilitate efficient 3D scene reconstruction without costly annotations, we propose a self-supervised street Gaussian ($\textit{S}^3$Gaussian) method to decompose dynamic and static elements from 4D consistency. We represent each scene with 3D Gaussians to preserve the explicitness and further accompany them with a spatial-temporal field network to compactly model the 4D dynamics. We conduct extensive experiments on the challenging Waymo-Open dataset to evaluate the effectiveness of our method. Our $\textit{S}^3$Gaussian demonstrates the ability to decompose static and dynamic scenes and achieves the best performance without using 3D annotations. Code is available at: https://github.com/nnanhuang/S3Gaussian/.