arXiv subjects

Tianchen Guo

Publications and source records attributed to Tianchen Guo.

5 recordsLinked to original sources

arXiv2026

When Depth Lies: Benchmarking Vision-Language Models on Mirror-Induced RGB-D Ambiguity

Vision-language models (VLMs) have made significant progress in spatial reasoning, particularly by integrating RGB images with depth data, which is often assumed to faithfully reflect physical distances. However, mirrors violate this assumption: they distort depth readings while reproducing object appearances, making reflections difficult to distinguish from real objects. As existing benchmarks rarely consider such reflective scenarios, VLM performance in mirror scenes remains largely unexplored. In this work, we introduce a mirror-centric RGB-D dataset from reflective scenes to evaluate the spatial reasoning capabilities of state-of-the-art VLMs in these scenarios. Specifically, we collect 1,011 RGB-D images from 525 real-world scenes, spanning diverse indoor and outdoor environments, six reflective surface types, and both daytime and nighttime conditions. All images are annotated with reflective surface region masks and mask annotations for both directly observed and mirror-reflected objects. Then, we curate a benchmark named MIRA-Bench (Mirror-Induced Reasoning Ambiguity Benchmark), comprising 14,140 QA pairs across eight sub-tasks organized into three cognitive levels: Reflection-aware Perception, Spatial Reasoning, and Scene-level Decision-making. We evaluate state-of-the-art VLMs on MIRA-Bench and find consistent failures across tasks: models misclassify whether objects are directly observed or reflected, fail to match reflected instances to their real-world counterparts, and produce erroneous spatial inferences. Remarkably, providing depth data actually degrades reasoning performance, implying that VLMs cannot identify incorrect depth in mirror scenes. The code and data are available at https://github.com/EvaYin-cloud/MIRA-Bench

cs.CV↗

arXiv2026

Humans in Space: Can MLLMs Construct Human-Centric Spatial Understanding via Active Exocentric Exploration?

While Multimodal Large Language Models (MLLMs) excel at zero-shot reasoning, their spatial capabilities are primarily evaluated through static, passive observation. In complex, human-centric scenes, occlusion often hides crucial geometric evidence, and passive benchmarks force models to guess rather than recognize their own visual uncertainty. To bridge this gap, we introduce SATE-Bench, a novel interactive benchmark comprising 316 real-world scenes and 2,525 multi-turn QA pairs to evaluate visual self-awareness and active exocentric view-exploration. It combines a Passive multi-view control with an Active exploration setting, where models start from a severely occluded view and must iteratively query adjacent viewpoints until they gather enough evidence to fully ``sate'' the question, scored by a step-penalized efficiency reward. Evaluating state-of-the-art MLLMs reveals that they terminate exploration prematurely while reporting high confidence, and struggle with fine-grained 3D geometric tasks; matched-view controls further attribute the Active--Passive gap primarily to insufficient evidence acquisition. Moreover, fine-tuning with Group Relative Policy Optimization (GRPO) shows that rewarding view-seeking increases active exploration but lowers overall accuracy, as models over-explore and fail to fuse expanding multi-view contexts across turns. These findings indicate that scaling passive multi-view perception does not intrinsically yield active spatial reasoning, highlighting the need for models that know where to look, how to integrate what they see, and when to stop. Code and data are available at https://gtc-gh.github.io/SATE-Bench/

cs.CV↗

arXiv2026

SSMNBench: Diagnosing Image-based Cross-View Human-Object Understanding via Single-View Sufficiency and Multi-View Necessity

Multimodal Large Language Models (MLLMs) have shown remarkable progress in single-image perception, yet their ability to reason about complex cross-view human-centric scenes remains largely unverified. Current multi-view benchmarks evaluate models using a fixed "bag of frames" and thus conflate a model's robustness to visual distraction with its genuine ability to fuse fragmented cross-view evidence. To address this issue, we introduce SSMNBench, a diagnostic benchmark comprising 3,300 curated QA pairs for cross-view human and human-object understanding. SSMNBench uniquely categorizes tasks into Single-View Sufficiency (SVS) and Multi-View Necessity (MVN). By systematically perturbing view availability across 17 state-of-the-art MLLMs, critical limitations are revealed: models suffer from severe "distraction degradation" when presented with redundant views (SVS), and fail to integrate fragmented geometric evidence across cameras (MVN). Our evaluations demonstrate that modern MLLMs rely on multiple single-image semantic averaging and view preference rather than genuine cross-view synthesis. By exposing these fundamental vulnerabilities, SSMNBench provides a rigorous diagnostic framework to drive the advancement of future cross-view-aware multimodal architectures. The code is available at: $ \href{https://github.com/gtc-gh/SSMNBench}{\text{SSMNBench}} $

cs.CV↗

arXiv2025

An Interpretable AI framework Quantifying Traditional Chinese Medicine Principles Towards Enhancing and Integrating with Modern Biomedicine

Traditional Chinese Medicine diagnosis and treatment principles, established through centuries of trial-and-error clinical practice, directly maps patient-specific symptom patterns to personalised herbal therapies. These empirical holistic mapping principles offer valuable strategies to address remaining challenges of reductionism methodologies in modern biomedicine. However, the lack of a quantitative framework and molecular-level evidence has limited their interpretability and reliability. Here, we present an AI framework trained on ancient and classical TCM formula records to quantify the symptom pattern-herbal therapy mappings. Interestingly, we find that empirical TCM diagnosis and treatment are consistent with the encoding-decoding processes in the AI model. This enables us to construct an interpretable TCM embedding space (TCM-ES) using the model's quantitative representation of TCM principles. Validated through broad and extensive TCM patient data, the TCM-ES offers universal quantification of the TCM practice and therapeutic efficacy. We further map biomedical entities into the TCM-ES through correspondence alignment. We find that the principal directions of the TCM-ES are significantly associated with key biological functions (such as metabolism, immune, and homeostasis), and that the disease and herb embedding proximity aligns with their genetic relationships in the human protein interactome, which demonstrate the biological significance of TCM principles. Moreover, the TCM-ES uncovers latent disease relationships, and provides alternative metric to assess clinical efficacy for modern disease-drug pairs. Finally, we construct a comprehensive and integrative TCM knowledge graph, which predicts potential associations between diseases and targets, drugs, herbal compounds, and herbal therapies, providing TCM-informed opportunities for disease analysis and drug development.

q-bio.OT↗

arXiv2024

A Large-scale Investigation of Semantically Incompatible APIs behind Compatibility Issues in Android Apps

Application Programming Interface (API) incompatibility is a long-standing issue in Android application development. The rapid evolution of Android APIs results in a significant number of API additions, removals, and changes between adjacent versions. Unfortunately, this high frequency of alterations may lead to compatibility issues, often without adequate notification to developers regarding these changes. Although researchers have proposed some work on detecting compatibility issues caused by changes in API signatures, they often overlook compatibility issues stemming from sophisticated semantic changes. In response to this challenge, we conducted a large-scale discovery of incompatible APIs in the Android Open Source Project (AOSP) by leveraging static analysis and pre-trained Large Language Models (LLMs) across adjacent versions. We systematically formulate the problem and propose a unified framework to detect incompatible APIs, especially for semantic changes. It's worth highlighting that our approach achieves a 0.83 F1-score in identifying semantically incompatible APIs in the Android framework. Ultimately, our approach detects 5,481 incompatible APIs spanning from version 4 to version 33. We further demonstrate its effectiveness in supplementing the state-of-the-art methods in detecting a broader spectrum of compatibility issues (+92.3%) that have been previously overlooked.

cs.SE↗