Searcharxiv⌕ Search

arXiv subjects

Jihan Yang

Publications and source records attributed to Jihan Yang.

At least 19 recordsLinked to original sources
arXiv2026

Cambrian-P: Pose-Grounded Video Understanding

Camera pose matters. The position and orientation of each viewpoint define a shared spatial coordinate frame that relates observations across video frames. Yet this signal is largely absent from multimodal LLMs (MLLMs) for video understanding, which process frames as isolated 2D snapshots, instead of the persistent scene humans perceive. We revisit pose as a lightweight supervisory signal and introduce Cambrian-P, a video MLLM augmented with per-frame learnable camera tokens and a pose regression head. With a carefully designed sampling scheme, the model achieves substantial gains of 4.5-6.5% on spatial reasoning benchmarks such as VSI-Bench, generalizes across eight additional spatial and general video QA benchmarks, and, as a byproduct, achieves state of the art streaming pose estimation on ScanNet. Surprisingly, training on pseudo-annotated poses from in-the-wild video further improves general video QA benchmarks, showing pose helps beyond spatial reasoning. Together, these results position camera pose as a fundamental signal for video models that reason about the physical world.

cs.CV↗
arXiv2026

From Where Things Are to What They Are For: Benchmarking Spatial-Functional Intelligence in Multimodal LLMs

Human-level agentic intelligence extends beyond low-level geometric perception, evolving from recognizing where things are to understanding what they are for. While existing benchmarks effectively evaluate the geometric perception capabilities of multimodal large language models (MLLMs), they fall short of probing the higher-order cognitive abilities required for grounded intelligence. To address this gap, we introduce the Spatial-Functional Intelligence Benchmark (SFI-Bench), a video-based benchmark with over 1,500 expert-annotated questions derived from diverse egocentric indoor video scans. SFI-Bench systematically evaluates two complementary dimensions of advanced reasoning: (1) Structured Spatial Reasoning, which requires understanding complex layouts and forming coherent spatial representations, and (2) Functional Reasoning, which involves inferring object affordances and their context-dependent utility. The benchmark includes tasks such as conditional counting, multi-hop relational reasoning, functional pairing, and knowledge-grounded troubleshooting, directly challenging models to integrate perception, memory, and inference. Our experiments reveal that current MLLMs consistently struggle to combine spatial memory with functional reasoning and external knowledge, highlighting a critical bottleneck in achieving grounded intelligence. SFI-Bench therefore provides a diagnostic tool for measuring progress toward more cognitively capable and truly grounded multimodal agents.

cs.CV↗
arXiv2026

Scaling Text-to-Image Diffusion Transformers with Representation Autoencoders

Representation Autoencoders (RAEs) have shown distinct advantages in diffusion modeling on ImageNet by training in high-dimensional semantic latent spaces. In this work, we investigate whether this framework can scale to large-scale, freeform text-to-image (T2I) generation. We first scale RAE decoders on the frozen representation encoder (SigLIP-2) beyond ImageNet by training on web, synthetic, and text-rendering data, finding that while scale improves general fidelity, targeted data composition is essential for specific domains like text. We then rigorously stress-test the RAE design choices originally proposed for ImageNet. Our analysis reveals that scaling simplifies the framework: while dimension-dependent noise scheduling remains critical, architectural complexities such as wide diffusion heads and noise-augmented decoding offer negligible benefits at scale Building on this simplified framework, we conduct a controlled comparison of RAE against the state-of-the-art FLUX VAE across diffusion transformer scales from 0.5B to 9.8B parameters. RAEs consistently outperform VAEs during pretraining across all model scales. Further, during finetuning on high-quality datasets, VAE-based models catastrophically overfit after 64 epochs, while RAE models remain stable through 256 epochs and achieve consistently better performance. Across all experiments, RAE-based diffusion models demonstrate faster convergence and better generation quality, establishing RAEs as a simpler and stronger foundation than VAEs for large-scale T2I generation. Additionally, because both visual understanding and generation can operate in a shared representation space, the multimodal model can directly reason over generated latents, opening new possibilities for unified models.

cs.CV↗
arXiv2025

Benchmark Designers Should "Train on the Test Set" to Expose Exploitable Non-Visual Shortcuts

Robust benchmarks are crucial for evaluating Multimodal Large Language Models (MLLMs). Yet we find that models can ace many multimodal benchmarks without strong visual understanding, instead exploiting biases, linguistic priors, and superficial patterns. This is especially problematic for vision-centric benchmarks that are meant to require visual inputs. We adopt a diagnostic principle for benchmark design: if a benchmark can be gamed, it will be. Designers should therefore try to ``game'' their own benchmarks first, using diagnostic and debiasing procedures to systematically identify and mitigate non-visual biases. Effective diagnosis requires directly ``training on the test set'' -- probing the released test set for its intrinsic, exploitable patterns. We operationalize this standard with two components. First, we diagnose benchmark susceptibility using a ``Test-set Stress-Test'' (TsT) methodology. Our primary diagnostic tool involves fine-tuning a powerful Large Language Model via $k$-fold cross-validation on exclusively the non-visual, textual inputs of the test set to reveal shortcut performance and assign each sample a bias score $s(x)$. We complement this with a lightweight Random Forest-based diagnostic operating on hand-crafted features for fast, interpretable auditing. Second, we debias benchmarks by filtering high-bias samples using an ``Iterative Bias Pruning'' (IBP) procedure. Applying this framework to four benchmarks -- VSI-Bench, CV-Bench, MMMU, and VideoMME -- we uncover pervasive non-visual biases. As a case study, we apply our full framework to create VSI-Bench-Debiased, demonstrating reduced non-visual solvability and a wider vision-blind performance gap than the original.

cs.CV↗
arXiv2025

Cambrian-S: Towards Spatial Supersensing in Video

We argue that progress in true multimodal intelligence calls for a shift from reactive, task-driven systems and brute-force long context towards a broader paradigm of supersensing. We frame spatial supersensing as four stages beyond linguistic-only understanding: semantic perception (naming what is seen), streaming event cognition (maintaining memory across continuous experiences), implicit 3D spatial cognition (inferring the world behind pixels), and predictive world modeling (creating internal models that filter and organize information). Current benchmarks largely test only the early stages, offering narrow coverage of spatial cognition and rarely challenging models in ways that require true world modeling. To drive progress in spatial supersensing, we present VSI-SUPER, a two-part benchmark: VSR (long-horizon visual spatial recall) and VSC (continual visual spatial counting). These tasks require arbitrarily long video inputs yet are resistant to brute-force context expansion. We then test data scaling limits by curating VSI-590K and training Cambrian-S, achieving +30% absolute improvement on VSI-Bench without sacrificing general capabilities. Yet performance on VSI-SUPER remains limited, indicating that scale alone is insufficient for spatial supersensing. We propose predictive sensing as a path forward, presenting a proof-of-concept in which a self-supervised next-latent-frame predictor leverages surprise (prediction error) to drive memory and event segmentation. On VSI-SUPER, this approach substantially outperforms leading proprietary baselines, showing that spatial supersensing requires models that not only see but also anticipate, select, and organize experience.

cs.CV↗
arXiv2025

Traveling Across Languages: Benchmarking Cross-Lingual Consistency in Multimodal LLMs

The rapid evolution of multimodal large language models (MLLMs) has significantly enhanced their real-world applications. However, achieving consistent performance across languages, especially when integrating cultural knowledge, remains a significant challenge. To better assess this issue, we introduce two new benchmarks: KnowRecall and VisRecall, which evaluate cross-lingual consistency in MLLMs. KnowRecall is a visual question answering benchmark designed to measure factual knowledge consistency in 15 languages, focusing on cultural and historical questions about global landmarks. VisRecall assesses visual memory consistency by asking models to describe landmark appearances in 9 languages without access to images. Experimental results reveal that state-of-the-art MLLMs, including proprietary ones, still struggle to achieve cross-lingual consistency. This underscores the need for more robust approaches that produce truly multilingual and culturally aware models.

cs.CL↗
arXiv2025

UniTok: A Unified Tokenizer for Visual Generation and Understanding

Visual generative and understanding models typically rely on distinct tokenizers to process images, presenting a key challenge for unifying them within a single framework. Recent studies attempt to address this by connecting the training of VQVAE (for autoregressive generation) and CLIP (for understanding) to build a unified tokenizer. However, directly combining these training objectives has been observed to cause severe loss conflicts. In this paper, we show that reconstruction and semantic supervision do not inherently conflict. Instead, the underlying bottleneck stems from limited representational capacity of discrete token space. Building on these insights, we introduce UniTok, a unified tokenizer featuring a novel multi-codebook quantization mechanism that effectively scales up the vocabulary size and bottleneck dimension. In terms of final performance, UniTok sets a new record of 0.38 rFID and 78.6% zero-shot accuracy on ImageNet. Besides, UniTok can be seamlessly integrated into MLLMs to unlock native visual generation capability, without compromising the understanding performance. Additionally, we show that UniTok favors cfg-free generation, reducing gFID from 14.6 to 2.5 on ImageNet 256$\times$256 benchmark. GitHub: https://github.com/FoundationVision/UniTok.

cs.CV↗
arXiv2025

SFT Memorizes, RL Generalizes: A Comparative Study of Foundation Model Post-training

Supervised fine-tuning (SFT) and reinforcement learning (RL) are widely used post-training techniques for foundation models. However, their roles in enhancing model generalization capabilities remain unclear. This paper studies the difference between SFT and RL on generalization and memorization, focusing on text-based rule variants and visual variants. We introduce GeneralPoints, an arithmetic reasoning card game, and adopt V-IRL, a real-world navigation environment, to assess how models trained with SFT and RL generalize to unseen variants in both textual and visual domains. We show that RL, especially when trained with an outcome-based reward, generalizes across both rule-based textual and visual variants. SFT, in contrast, tends to memorize training data and struggles to generalize out-of-distribution scenarios. Further analysis reveals that RL improves the model's underlying visual recognition capabilities, contributing to its enhanced generalization in the visual domain. Despite RL's superior generalization, we show that SFT remains essential for effective RL training; SFT stabilizes the model's output format, enabling subsequent RL to achieve its performance gains. These findings demonstrates the capability of RL for acquiring generalizable knowledge in complex, multi-modal tasks.

cs.AI↗
arXiv2024

Thinking in Space: How Multimodal Large Language Models See, Remember, and Recall Spaces

Humans possess the visual-spatial intelligence to remember spaces from sequential visual observations. However, can Multimodal Large Language Models (MLLMs) trained on million-scale video datasets also ``think in space'' from videos? We present a novel video-based visual-spatial intelligence benchmark (VSI-Bench) of over 5,000 question-answer pairs, and find that MLLMs exhibit competitive - though subhuman - visual-spatial intelligence. We probe models to express how they think in space both linguistically and visually and find that while spatial reasoning capabilities remain the primary bottleneck for MLLMs to reach higher benchmark performance, local world models and spatial awareness do emerge within these models. Notably, prevailing linguistic reasoning techniques (e.g., chain-of-thought, self-consistency, tree-of-thoughts) fail to improve performance, whereas explicitly generating cognitive maps during question-answering enhances MLLMs' spatial distance ability.

cs.CV↗
arXiv2024

Cambrian-1: A Fully Open, Vision-Centric Exploration of Multimodal LLMs

We introduce Cambrian-1, a family of multimodal LLMs (MLLMs) designed with a vision-centric approach. While stronger language models can enhance multimodal capabilities, the design choices for vision components are often insufficiently explored and disconnected from visual representation learning research. This gap hinders accurate sensory grounding in real-world scenarios. Our study uses LLMs and visual instruction tuning as an interface to evaluate various visual representations, offering new insights into different models and architectures -- self-supervised, strongly supervised, or combinations thereof -- based on experiments with over 20 vision encoders. We critically examine existing MLLM benchmarks, address the difficulties involved in consolidating and interpreting results from various tasks, and introduce a new vision-centric benchmark, CV-Bench. To further improve visual grounding, we propose the Spatial Vision Aggregator (SVA), a dynamic and spatially-aware connector that integrates high-resolution vision features with LLMs while reducing the number of tokens. Additionally, we discuss the curation of high-quality visual instruction-tuning data from publicly available sources, emphasizing the importance of data source balancing and distribution ratio. Collectively, Cambrian-1 not only achieves state-of-the-art performance but also serves as a comprehensive, open cookbook for instruction-tuned MLLMs. We provide model weights, code, supporting tools, datasets, and detailed instruction-tuning and evaluation recipes. We hope our release will inspire and accelerate advancements in multimodal systems and visual representation learning.

cs.CV↗
arXiv2024

RegionPLC: Regional Point-Language Contrastive Learning for Open-World 3D Scene Understanding

We propose a lightweight and scalable Regional Point-Language Contrastive learning framework, namely \textbf{RegionPLC}, for open-world 3D scene understanding, aiming to identify and recognize open-set objects and categories. Specifically, based on our empirical studies, we introduce a 3D-aware SFusion strategy that fuses 3D vision-language pairs derived from multiple 2D foundation models, yielding high-quality, dense region-level language descriptions without human 3D annotations. Subsequently, we devise a region-aware point-discriminative contrastive learning objective to enable robust and effective 3D learning from dense regional language supervision. We carry out extensive experiments on ScanNet, ScanNet200, and nuScenes datasets, and our model outperforms prior 3D open-world scene understanding approaches by an average of 17.2\% and 9.1\% for semantic and instance segmentation, respectively, while maintaining greater scalability and lower resource demands. Furthermore, our method has the flexibility to be effortlessly integrated with language models to enable open-ended grounded 3D reasoning without extra task-specific training. Code is available at https://github.com/CVMI-Lab/PLA.

cs.CV↗
arXiv2024

Can 3D Vision-Language Models Truly Understand Natural Language?

Rapid advancements in 3D vision-language (3D-VL) tasks have opened up new avenues for human interaction with embodied agents or robots using natural language. Despite this progress, we find a notable limitation: existing 3D-VL models exhibit sensitivity to the styles of language input, struggling to understand sentences with the same semantic meaning but written in different variants. This observation raises a critical question: Can 3D vision-language models truly understand natural language? To test the language understandability of 3D-VL models, we first propose a language robustness task for systematically assessing 3D-VL models across various tasks, benchmarking their performance when presented with different language style variants. Importantly, these variants are commonly encountered in applications requiring direct interaction with humans, such as embodied robotics, given the diversity and unpredictability of human language. We propose a 3D Language Robustness Dataset, designed based on the characteristics of human language, to facilitate the systematic study of robustness. Our comprehensive evaluation uncovers a significant drop in the performance of all existing models across various 3D-VL tasks. Even the state-of-the-art 3D-LLM fails to understand some variants of the same sentences. Further in-depth analysis suggests that the existing models have a fragile and biased fusion module, which stems from the low diversity of the existing dataset. Finally, we propose a training-free module driven by LLM, which improves language robustness. Datasets and code will be available at github.

cs.CV↗
arXiv2024

V-IRL: Grounding Virtual Intelligence in Real Life

There is a sensory gulf between the Earth that humans inhabit and the digital realms in which modern AI agents are created. To develop AI agents that can sense, think, and act as flexibly as humans in real-world settings, it is imperative to bridge the realism gap between the digital and physical worlds. How can we embody agents in an environment as rich and diverse as the one we inhabit, without the constraints imposed by real hardware and control? Towards this end, we introduce V-IRL: a platform that enables agents to scalably interact with the real world in a virtual yet realistic environment. Our platform serves as a playground for developing agents that can accomplish various practical tasks and as a vast testbed for measuring progress in capabilities spanning perception, decision-making, and interaction with real-world data across the entire globe.

cs.AI↗
arXiv2023

Lowis3D: Language-Driven Open-World Instance-Level 3D Scene Understanding

Open-world instance-level scene understanding aims to locate and recognize unseen object categories that are not present in the annotated dataset. This task is challenging because the model needs to both localize novel 3D objects and infer their semantic categories. A key factor for the recent progress in 2D open-world perception is the availability of large-scale image-text pairs from the Internet, which cover a wide range of vocabulary concepts. However, this success is hard to replicate in 3D scenarios due to the scarcity of 3D-text pairs. To address this challenge, we propose to harness pre-trained vision-language (VL) foundation models that encode extensive knowledge from image-text pairs to generate captions for multi-view images of 3D scenes. This allows us to establish explicit associations between 3D shapes and semantic-rich captions. Moreover, to enhance the fine-grained visual-semantic representation learning from captions for object-level categorization, we design hierarchical point-caption association methods to learn semantic-aware embeddings that exploit the 3D geometry between 3D points and multi-view images. In addition, to tackle the localization challenge for novel classes in the open-world setting, we develop debiased instance localization, which involves training object grouping modules on unlabeled data using instance-level pseudo supervision. This significantly improves the generalization capabilities of instance grouping and thus the ability to accurately locate novel objects. We conduct extensive experiments on 3D semantic, instance, and panoptic segmentation tasks, covering indoor and outdoor scenes across three datasets. Our method outperforms baseline methods by a significant margin in semantic segmentation (e.g. 34.5%$\sim$65.3%), instance segmentation (e.g. 21.8%$\sim$54.0%) and panoptic segmentation (e.g. 14.7%$\sim$43.3%). Code will be available.

cs.CV↗
arXiv2023

PLA: Language-Driven Open-Vocabulary 3D Scene Understanding

Open-vocabulary scene understanding aims to localize and recognize unseen categories beyond the annotated label space. The recent breakthrough of 2D open-vocabulary perception is largely driven by Internet-scale paired image-text data with rich vocabulary concepts. However, this success cannot be directly transferred to 3D scenarios due to the inaccessibility of large-scale 3D-text pairs. To this end, we propose to distill knowledge encoded in pre-trained vision-language (VL) foundation models through captioning multi-view images from 3D, which allows explicitly associating 3D and semantic-rich captions. Further, to foster coarse-to-fine visual-semantic representation learning from captions, we design hierarchical 3D-caption pairs, leveraging geometric constraints between 3D scenes and multi-view images. Finally, by employing contrastive learning, the model learns language-aware embeddings that connect 3D and text for open-vocabulary tasks. Our method not only remarkably outperforms baseline methods by 25.8% $\sim$ 44.7% hIoU and 14.5% $\sim$ 50.4% hAP$_{50}$ in open-vocabulary semantic and instance segmentation, but also shows robust transferability on challenging zero-shot domain transfer tasks. See the project website at https://dingry.github.io/projects/PLA.

cs.CV↗
arXiv2022

Towards Efficient 3D Object Detection with Knowledge Distillation

Despite substantial progress in 3D object detection, advanced 3D detectors often suffer from heavy computation overheads. To this end, we explore the potential of knowledge distillation (KD) for developing efficient 3D object detectors, focusing on popular pillar- and voxel-based detectors.In the absence of well-developed teacher-student pairs, we first study how to obtain student models with good trade offs between accuracy and efficiency from the perspectives of model compression and input resolution reduction. Then, we build a benchmark to assess existing KD methods developed in the 2D domain for 3D object detection upon six well-constructed teacher-student pairs. Further, we propose an improved KD pipeline incorporating an enhanced logit KD method that performs KD on only a few pivotal positions determined by teacher classification response, and a teacher-guided student model initialization to facilitate transferring teacher model's feature extraction ability to students through weight inheritance. Finally, we conduct extensive experiments on the Waymo dataset. Our best performing model achieves $65.75\%$ LEVEL 2 mAPH, surpassing its teacher model and requiring only $44\%$ of teacher flops. Our most efficient model runs 51 FPS on an NVIDIA A100, which is $2.2\times$ faster than PointPillar with even higher accuracy. Code is available at \url{https://github.com/CVMI-Lab/SparseKD}.

cs.CV↗
arXiv2022

DODA: Data-oriented Sim-to-Real Domain Adaptation for 3D Semantic Segmentation

Deep learning approaches achieve prominent success in 3D semantic segmentation. However, collecting densely annotated real-world 3D datasets is extremely time-consuming and expensive. Training models on synthetic data and generalizing on real-world scenarios becomes an appealing alternative, but unfortunately suffers from notorious domain shifts. In this work, we propose a Data-Oriented Domain Adaptation (DODA) framework to mitigate pattern and context gaps caused by different sensing mechanisms and layout placements across domains. Our DODA encompasses virtual scan simulation to imitate real-world point cloud patterns and tail-aware cuboid mixing to alleviate the interior context gap with a cuboid-based intermediate domain. The first unsupervised sim-to-real adaptation benchmark on 3D indoor semantic segmentation is also built on 3D-FRONT, ScanNet and S3DIS along with 7 popular Unsupervised Domain Adaptation (UDA) methods. Our DODA surpasses existing UDA approaches by over 13% on both 3D-FRONT -> ScanNet and 3D-FRONT -> S3DIS. Code is available at https://github.com/CVMI-Lab/DODA.

cs.CV↗
arXiv2022

Knowledge Distillation as Efficient Pre-training: Faster Convergence, Higher Data-efficiency, and Better Transferability

Large-scale pre-training has been proven to be crucial for various computer vision tasks. However, with the increase of pre-training data amount, model architecture amount, and the private/inaccessible data, it is not very efficient or possible to pre-train all the model architectures on large-scale datasets. In this work, we investigate an alternative strategy for pre-training, namely Knowledge Distillation as Efficient Pre-training (KDEP), aiming to efficiently transfer the learned feature representation from existing pre-trained models to new student models for future downstream tasks. We observe that existing Knowledge Distillation (KD) methods are unsuitable towards pre-training since they normally distill the logits that are going to be discarded when transferred to downstream tasks. To resolve this problem, we propose a feature-based KD method with non-parametric feature dimension aligning. Notably, our method performs comparably with supervised pre-training counterparts in 3 downstream tasks and 9 downstream datasets requiring 10x less data and 5x less pre-training time. Code is available at https://github.com/CVMI-Lab/KDEP.

cs.CV↗