Searcharxiv⌕ Search

arXiv subjects

Jiankang Deng

Publications and source records attributed to Jiankang Deng.

At least 37 records · Page 2Linked to original sources
arXiv2026

Metis: A Generalizable and Efficient World-Action Model for Autonomous Driving and Urban Navigation

World action models~(WAMs) have shown great promise for autonomous driving and urban navigation. Built upon Vision-Language-Action models or video generation models, existing approaches suffer key limitations: (1) High inference latency due to future observation prediction at test time, and (2) tightly coupled video and action modeling leading to representational mismatch and degraded generalization. To address both issues, we propose Metis, an end-to-end WAM framework that decouples video generation and action prediction. Specifically, Metis employs a Mixture-of-Transformers architecture with dedicated experts for video generation and action prediction, preserving the intrinsic distributional properties of each task. To enhance efficiency, we introduce an asymmetric attention mask that enables joint training of both experts while allowing the action model to bypass explicit video generation during inference. This design ensures training-inference consistency and significantly reduces computational costs without compromising planning performance. Extensive experiments demonstrate state-of-the-art performance on the NAVSIM navhard and navtest benchmarks and the CityWalker navigation benchmark, validating both the generalizability and efficiency across diverse tasks. Real-robot deployments further confirm the practical feasibility of our approach.

cs.CV↗
arXiv2026

LiteVSR: Lightweight Adaptation of Frozen Diffusion Transformers for Video Super-Resolution

Adapting large-scale pre-trained video generators for Video Super-Resolution (VSR) in novel domains remains computationally prohibitive. Methods that reformulate generation as direct Low-Quality to High-Quality mappings deviate from the original generative formulation, demanding extensive fine-tuning. ControlNet-style adapters lose their efficiency under modern Diffusion Transformers since the absence of encoder-decoder hierarchy forces duplication of the entire backbone. We observe that flow matching offers a principled alternative for cross-domain VSR adaptation. By predicting a constant velocity field across all timesteps, the adaptation task reduces to learning a fixed injection pattern rather than time-varying transformations. Building on this insight, we propose LiteVSR, a minimalist framework that performs VSR using a completely frozen Diffusion Transformer with a lightweight State-Aware Adapter. The adapter employs a dual-stream architecture that extracts static structural cues from the LQ input and dynamic cues from intermediate denoising states, aligning them through time-dependent cross-attention to enable adaptive transition from structural alignment to texture refinement as denoising proceeds. LiteVSR achieves competitive restoration quality with only 11.25% trainable parameters and 12 GPU-hours of training on a single A100, while maintaining fast sampling (down to a single step) compatibility.

cs.CV↗
arXiv2026

CASteer: Cross-Attention Steering for Controllable Concept Erasure

Diffusion models have transformed image generation, yet controlling their outputs to reliably erase undesired concepts remains challenging. Existing approaches usually require task-specific training and struggle to generalize across both concrete (e.g., objects) and abstract (e.g., styles) concepts. We propose CASteer (Cross-Attention Steering), a training-free framework for concept erasure in diffusion models using steering vectors to influence hidden representations dynamically. CASteer precomputes concept-specific steering vectors by averaging neural activations from images generated for each target concept. During inference, it dynamically applies these vectors to suppress undesired concepts only when they appear, ensuring that unrelated regions remain unaffected. This selective activation enables precise, context-aware erasure without degrading overall image quality. This approach achieves effective removal of harmful or unwanted content across a wide range of visual concepts, all without model retraining. CASteer outperforms state-of-the-art concept erasure techniques while preserving unrelated content and minimizing unintended effects.

cs.GR↗
arXiv2026

MidSteer: Optimal Affine Framework for Steering Generative Models

Steering intermediate representations has emerged as a powerful strategy for controlling generative models, particularly in post-deployment alignment and safety settings. However, despite its empirical success, it currently lacks a comprehensive theoretical framework. In this paper, we bridge this gap by formalizing the theory of concept steering. First, we establish a link between steering and affine concept erasure, proving that the standard approach for removing unwanted behaviors is a special case of LEACE (a closed-form method for affine erasure). Next, we formulate a principled theoretical framework for concept switching, LEACE-Switch, and characterize the assumptions under which it provides an optimal affine solution. Building on this analysis, we then introduce MidSteer (Minimal Disturbance concept Steering), a more general affine framework for concept manipulation that relaxes these assumptions and enables directed, minimal-disturbance transformations. We demonstrate that MidSteer performs favorably across a range of tasks, modalities, and architectures, including vision diffusion models and large language models.

cs.LG↗
arXiv2026

VisualFLIP: Do Predictions Depend on Task-Critical Visual Evidence in Multimodal Reasoning?

When a multimodal large language model answers a visual reasoning question correctly, is the prediction actually supported by the task-critical visual evidence? Correct answers can coexist with flawed reasoning, making accuracy alone an incomplete test of grounding. We introduce VisualFLIP, a paired benchmark with 1,374 images arranged as same-question perturbation pairs across cardinality, attribute, spatial, and logic tasks. Each pair keeps the question fixed but minimally changes the evidence so the gold answer deterministically flips. We evaluate 24 MLLMs with pair accuracy, which requires solving both sides of a pair, and Collapse Rate (CR), which measures how often a model that solves at least one side repeats the same non-empty answer for both images. Together, these metrics show that paired correctness and evidence dependence are related but distinct: capable models can still fail to update after task-critical visual changes, and collapse becomes more severe for some models when the edited image follows an earlier answer in a sequential setting. Further details are available on our project page: https://didizhu-judy.github.io/VisualFLIP/

cs.CV↗
arXiv2026

Chirpy3D: Part-Aware Multi-View Diffusion for Creative Fine-Grained Object Generation

Understanding and generating the fine-grained structure of objects -- such as birds with species-specific beaks, wings, and tails -- is a long-standing challenge in computer vision. We propose Chirpy3D, a part-aware multi-view diffusion framework that learns a hierarchical part latent space from unposed 2D images, using only off-the-shelf 2D part segmentation masks as spatial guidance -- without requiring any 3D data, camera poses, or manual part annotations. This latent space enables intuitive part-level swapping, interpolation, and zero-shot composition. A self-supervised feature consistency loss further encourages structural alignment across views, allowing coherent generation even with hybrid or unseen part combinations. Our core contribution is the controllable part-aware latent space and multi-view diffusion model. Downstream 3D generation is supported via any differentiable renderer such as NeRF but is orthogonal to the main framework, making Chirpy3D a flexible foundation for creative object generation in the absence of structured 3D data. Code is released at https://github.com/kamwoh/chirpy3d.

cs.CV↗
arXiv2026

LLaVA-OneVision-2: Towards Next-Generation Perceptual Intelligence

We introduce LLaVA-OneVision-2 (LLaVA-OV-2), the most capable vision-language model in the LLaVA-OneVision series to date, achieving superior performance across a broad range of multimodal benchmarks. The model builds on a native OneVision-Encoder and incorporates Windowed Attention for efficient local computation while maintaining native resolution. Its key advance is codec-stream tokenization: it treats compressed video as a continuous bit-cost stream, where bit-cost dynamics determine adaptive temporal groups, and motion-residual cues select salient spatial evidence into compact visual canvases. This allocation concentrates a limited token budget on event-bearing content, enabling more stable long-video token compression than fixed groups of pictures. A shared 3D RoPE further places codec canvases, sampled frames, and images in a unified spatiotemporal coordinate system. Furthermore, we build the LLaVA-OV-2 data and training stack around large-scale open supervision: approximately 8M re-captioned video samples for pretraining, a 4M-sample spatial corpus for fine-tuning. We also introduce JumpScore, a temporal-localization benchmark targeting fine-grained grounding in high-frequency, densely repeated motion, a regime underrepresented by existing video evaluations. A standout capability of LLaVA-OV-2 is its unified perception across video understanding, temporal grounding, spatial grounding, and manipulation-trace reasoning. On JumpScore, LLaVA-OneVision-2-8B reaches 74.9 JumpScore mAP, surpassing Qwen3-VL-8B (30.1) by +44.8 points; under matched visual-token budgets on the same benchmark, codec-stream inputs improve temporal grounding over frame sampling by +9.7 points. Across standard benchmarks, LLaVA-OneVision-2-8B further outperforms Qwen3-VL-8B by +4.3 average points on video tasks, +5.3 on spatial tasks, and +15.6 average J&F on tracking tasks.

cs.CV↗
arXiv2026

RAVEN: Real-time Autoregressive Video Extrapolation with Consistency-model GRPO

Causal autoregressive video diffusion models support real-time streaming generation by extrapolating future chunks from previously generated content. Distilling such generators from high-fidelity bidirectional teachers yields competitive few-step models, yet a persistent gap between the history distributions encountered during training and those arising at inference constrains generation quality over long horizons. We introduce the Real-time Autoregressive Video Extrapolation Network (RAVEN), a training-time test framework that repacks each self rollout into an interleaved sequence of clean historical endpoints and noisy denoising states. This formulation aligns training attention with inference-time extrapolation and allows downstream chunk losses to supervise the history representations on which future predictions depend. We further propose Consistency-model Group Relative Policy Optimization (CM-GRPO), which reformulates a consistency sampling step as a conditional Gaussian transition and applies online Reinforcement Learning (RL) directly to this kernel, avoiding the Euler-Maruyama auxiliary process adopted in prior flow-model RL formulations. Experiments demonstrate that RAVEN surpasses recent causal video distillation baselines across quality, semantic, and dynamic degree evaluations, and that CM-GRPO provides further gains when combined with RAVEN.

cs.CV↗
arXiv2026

From Imagined Futures to Executable Actions: Mixture of Latent Actions for Robot Manipulation

Video generation models offer a promising imagination mechanism for robot manipulation by predicting long-horizon future observations, but effectively exploiting these imagined futures for action execution remains challenging. Existing approaches either condition policies on predicted frames or directly decode generated videos into actions, both suffering from a mismatch between visual realism and control relevance. As a result, predicted observations emphasize perceptual fidelity rather than action-centric causes of state transitions, leading to indirect and unstable control. To address this gap, we propose MoLA (Mixture of Latent Actions), a control-oriented interface that transforms imagined future videos into executable representations. Instead of passing predicted frames directly to the policy, MoLA leverages a mixture of pretrained inverse dynamics models to infer a mixture of latent actions implied by generated visual transitions. These modality-aware inverse dynamics models capture complementary semantic, depth, and flow cues, providing a structured and physically grounded action representation that bridges video imagination and policy execution. We evaluate our approach on simulated benchmarks (LIBERO, CALVIN, and LIBERO-Plus) and real-world robot manipulation tasks, achieving consistent gains in task success, temporal consistency, and generalization.

cs.RO↗
arXiv2026

See Tomorrow, Act Today: Foresight-Driven Autonomous Driving

Current end-to-end autonomous driving planners are fundamentally reactive: they condition on historical and present observations to predict future actions. We argue that autonomous agents should instead imagine future scenes before deciding, just as human drivers mentally simulate ``what will happen next" before acting. We introduce ForeSight, a foundation world model centric planning framework that reframes autonomous driving as anticipatory decision-making. Rather than treating world models as auxiliary components, ForeSight makes future scene imagination the primary driver of action prediction. Our approach operates in two stages: (1) generating plausible future visual worlds via a pretrained world model, and (2) planning actions conditioned on these imagined futures. This paradigm shift from ``what should I do now?" to ``what will happen, and how should I respond?" enables genuinely anticipatory rather than reactive planning. By grounding decisions in anticipated contexts rather than present observations alone, ForeSight navigates dynamic, interactive scenarios more effectively. Extensive experiments on NAVSIM and nuScenes demonstrate that explicit future imagination significantly outperforms previous state-of-the-art alternatives, validating our foresight-driven approach.

cs.CV↗
arXiv2026

MoCoTalk: Multi-Conditional Diffusion with Adaptive Router for Controllable Talking Head Generation

Talking-head generation requires joint modeling of identity, head pose, facial expression, and mouth dynamics. Existing methods typically address only a subset of these factors, and rely on fixed-weight or heuristic fusion when multiple conditions are involved. We present MoCoTalk, a multi-conditional video diffusion framework that unifies four complementary control signals: a reference image, facial keypoints, 3DMM-rendered shading meshes, and the corresponding speech audio. To resolve destructive interference among heterogeneous conditions, we introduce an Adaptive Multi-Condition Router that computes channel-wise, timestep-aware gating over the four condition streams, allowing the fusion strategy to vary with both feature subspace and noise level. To better capture speech-related facial dynamics, we design a Mouth-Augmented Shading Mesh, a 3DMM-based representation that decouples head motion, mouth motion, expression, and lighting. This design provides a temporally consistent geometric prior and allows flexible recombination of these attributes at inference. We further introduce a lip consistency loss to tighten audio-visual alignment. Extensive experiments show that MoCoTalk achieves state-of-the-art performance on the majority of structural, motion, and perceptual metrics, while offering attribute-level controllability that single-condition methods do not provide.

cs.CV↗
arXiv2026

High Dynamic Range 3D Gaussian Splatting via Luminance-Chromaticity Decomposition

High Dynamic Range (HDR) 3D reconstruction is pivotal for professional content creation in filmmaking and virtual production. Existing methods typically rely on multi-exposure Low Dynamic Range (LDR) supervision to constrain the learning process within vast brightness spaces, resulting in complex, dual-branch architectures. This work explores the feasibility of learning HDR 3D models exclusively in the HDR data space to simplify model design. By analyzing 3D Gaussian Splatting (3DGS) for HDR imagery, we reveal that its failure stems from the limited capacity of Spherical Harmonics (SHs) to capture extreme radiance variations across views, often biasing towards high-radiance observations and underfitting. While increasing the maximum SH degree improves training fitting, it leads to severe overfitting and excessive parameter overhead. To address this, we propose \textit{Luminance--Chromaticity Decomposition Gaussian Splatting} (LCD-GS). By decoupling luminance and chromaticity into independent parameters, LCD-GS significantly enhances learning flexibility with minimal parameter increase (\textit{e.g.}, one extra scalar per primitive). Notably, LCD-GS maintains the original training and inference pipeline, requiring only a change in color representation. This explicit decomposition naturally enables primitive-level local and global luminance editing during inference. Extensive experiments on synthetic and real datasets demonstrate that LCD-GS consistently outperforms state-of-the-art methods in reconstruction fidelity and dynamic-range preservation even with a simpler, more efficient architecture, providing an elegant paradigm for professional-grade HDR 3D modeling. Code and datasets will be released.

cs.CV↗
arXiv2026

FreeScale: Scaling 3D Scenes via Certainty-Aware Free-View Generation

The development of generalizable Novel View Synthesis (NVS) models is critically limited by the scarcity of large-scale training data featuring diverse and precise camera trajectories. While real-world captures are photorealistic, they are typically sparse and discrete. Conversely, synthetic data scales but suffers from a domain gap and often lacks realistic semantics. We introduce FreeScale, a novel framework that leverages the power of scene reconstruction to transform limited real-world image sequences into a scalable source of high-quality training data. Our key insight is that an imperfect reconstructed scene serves as a rich geometric proxy, but naively sampling from it amplifies artifacts. To this end, we propose a certainty-aware free-view sampling strategy identifying novel viewpoints that are both semantically meaningful and minimally affected by reconstruction errors. We demonstrate FreeScale's effectiveness by scaling up the training of feedforward NVS models, achieving a notable gain of 2.7 dB in PSNR on challenging out-of-distribution benchmarks. Furthermore, we show that the generated data can actively enhance per-scene 3D Gaussian Splatting optimization, leading to consistent improvements across multiple datasets. Our work provides a practical and powerful data generation engine to overcome a fundamental bottleneck in 3D vision. Project page: https://mvp-ai-lab.github.io/FreeScale.

cs.CV↗
arXiv2026

HO-Flow: Generalizable Hand-Object Interaction Generation with Latent Flow Matching

Generating realistic 3D hand-object interactions (HOI) is a fundamental challenge in computer vision and robotics, requiring both temporal coherence and high-fidelity physical plausibility. Existing methods remain limited in their ability to learn expressive motion representations for generation and perform temporal reasoning. In this paper, we present HO-Flow, a framework for synthesizing realistic hand-object motion sequences from texts and canoncial 3D objects. HO-Flow first employs an interaction-aware variational autoencoder to encode sequences of hand and object motions into a unified latent manifold by incorporating hand and object kinematics, enabling the representation to capture rich interaction dynamics. It then leverages a masked flow matching model that combines auto-regressive temporal reasoning with continuous latent generation, improving temporal coherence. To further enhance generalization, HO-Flow predicts object motions relative to the initial frame, enabling effective pre-training on large-scale synthetic data. Experiments on the GRAB, OakInk, and DexYCB benchmarks demonstrate that HO-Flow achieves state-of-the-art performance in both physical plausibility and motion diversity for interaction motion synthesis.

cs.CV↗
arXiv2026

PaCo-FR: Patch-Pixel Aligned End-to-End Codebook Learning for Facial Representation Pre-training

Facial representation pre-training is crucial for tasks like facial recognition, expression analysis, and virtual reality. However, existing methods face three key challenges: (1) failing to capture distinct facial features and fine-grained semantics, (2) ignoring the spatial structure inherent to facial anatomy, and (3) inefficiently utilizing limited labeled data. To overcome these, we introduce PaCo-FR, an unsupervised framework that combines masked image modeling with patch-pixel alignment. Our approach integrates three innovative components: (1) a structured masking strategy that preserves spatial coherence by aligning with semantically meaningful facial regions, (2) a novel patch-based codebook that enhances feature discrimination with multiple candidate tokens, and (3) spatial consistency constraints that preserve geometric relationships between facial components. PaCo-FR achieves state-of-the-art performance across several facial analysis tasks with just 2 million unlabeled images for pre-training. Our method demonstrates significant improvements, particularly in scenarios with varying poses, occlusions, and lighting conditions. We believe this work advances facial representation learning and offers a scalable, efficient solution that reduces reliance on expensive annotated datasets, driving more effective facial analysis systems.

cs.CV↗
arXiv2026

Interact2Ar: Full-Body Human-Human Interaction Generation via Autoregressive Diffusion Models

Generating realistic human-human interactions is a challenging task that requires not only high-quality individual body and hand motions, but also coherent coordination among all interactants. Due to limitations in available data and increased learning complexity, previous methods tend to ignore hand motions, limiting the realism and expressivity of the interactions. Additionally, current diffusion-based approaches generate entire motion sequences simultaneously, limiting their ability to capture the reactive and adaptive nature of human interactions. To address these limitations, we introduce Interact2Ar, the first end-to-end text-conditioned autoregressive diffusion model for generating full-body, human-human interactions. Interact2Ar incorporates detailed hand kinematics through dedicated parallel branches, enabling high-fidelity full-body generation. Furthermore, we introduce an autoregressive pipeline coupled with a novel memory technique that facilitates adaptation to the inherent variability of human interactions using efficient large context windows. The adaptability of our model enables a series of downstream applications, including temporal motion composition, real-time adaptation to disturbances, and extension beyond dyadic to multi-person scenarios. To validate the generated motions, we introduce a set of robust evaluators and extended metrics designed specifically for assessing full-body interactions. Through quantitative and qualitative experiments, we demonstrate the state-of-the-art performance of Interact2Ar.

cs.CV↗
arXiv2026

Do You See What I Am Pointing At? Gesture-Based Egocentric Video Question Answering

Understanding and answering questions based on a user's pointing gesture is essential for next-generation egocentric AI assistants. However, current Multimodal Large Language Models (MLLMs) struggle with such tasks due to the lack of gesture-rich data and their limited ability to infer fine-grained pointing intent from egocentric video. To address this, we introduce EgoPointVQA, a dataset and benchmark for gesture-grounded egocentric question answering, comprising 4000 synthetic and 400 real-world videos across multiple deictic reasoning tasks. Built upon it, we further propose Hand Intent Tokens (HINT), which encodes tokens derived from 3D hand keypoints using an off-the-shelf reconstruction model and interleaves them with the model input to provide explicit spatial and temporal context for interpreting pointing intent. We show that our model outperforms others in different backbones and model sizes. In particular, HINT-14B achieves 68.1% accuracy, on average over 6 tasks, surpassing the state-of-the-art, InternVL3-14B, by 6.6%. To further facilitate the open research, we will release the code, model, and dataset. Project page: https://yuuraa.github.io/papers/choi2026egovqa

cs.CV↗
arXiv2026

Unleashing Vision-Language Semantics for Deepfake Video Detection

Recent Deepfake Video Detection (DFD) studies have demonstrated that pre-trained Vision-Language Models (VLMs) such as CLIP exhibit strong generalization capabilities in detecting artifacts across different identities. However, existing approaches focus on leveraging visual features only, overlooking their most distinctive strength -- the rich vision-language semantics embedded in the latent space. We propose VLAForge, a novel DFD framework that unleashes the potential of such cross-modal semantics to enhance model's discriminability in deepfake detection. This work i) enhances the visual perception of VLM through a ForgePerceiver, which acts as an independent learner to capture diverse, subtle forgery cues both granularly and holistically, while preserving the pretrained Vision-Language Alignment (VLA) knowledge, and ii) provides a complementary discriminative cue -- Identity-Aware VLA score, derived by coupling cross-modal semantics with the forgery cues learned by ForgePerceiver. Notably, the VLA score is augmented by an identity prior-informed text prompting to capture authenticity cues tailored to each identity, thereby enabling more discriminative cross-modal semantics. Comprehensive experiments on video DFD benchmarks, including classical face-swapping forgeries and recent full-face generation forgeries, demonstrate that our VLAForge substantially outperforms state-of-the-art methods at both frame and video levels. Code is available at https://github.com/mala-lab/VLAForge.

cs.CV↗