Lan Hu

Jiaxin Wei, Lan Hu, Chenyu Wang et al.

We present a new solution to the fine-grained retrieval of clean CAD models from a large-scale database in order to recover detailed object shape geometries for RGBD scans. Unlike previous work simply indexing into a moderately small database using an object shape descriptor and accepting the top retrieval result, we argue that in the case of a large-scale database a more accurate model may be found within a neighborhood of the descriptor. More importantly, we propose that the distinctiveness deficiency of shape descriptors at the instance level can be compensated by a geometry-based re-ranking of its neighborhood. Our approach first leverages the discriminative power of learned representations to distinguish between different categories of models and then uses a novel robust point set distance metric to re-rank the CAD neighborhood, enabling fine-grained retrieval in a large shape database. Evaluation on a real-world dataset shows that our geometry-based re-ranking is a conceptually simple but highly effective method that can lead to a significant improvement in retrieval accuracy compared to the state-of-the-art.

Binqi Shen, Lier Jin, Hanyu Cai et al.

Large language models (LLMs) increasingly rely on long-context processing, but expanding context windows introduces substantial computational and financial costs. Existing context reduction approaches, including retrieval and memory compression methods, are typically evaluated using performance and efficiency metrics independently, limiting systematic comparison and deployment-aware decision-making. This paper introduces The Efficiency Frontier, a unified framework for cost-performance optimization in LLM context management. The framework models context strategy selection as a deployment-aware optimization problem that jointly accounts for task performance, token cost, and preprocessing reuse through amortized cost modeling. Unlike existing evaluations that compare methods in isolation, the proposed framework enables decision-oriented analysis of when different context management strategies become preferable under varying operational conditions. Evaluated on 5,000 HotpotQA instances, the framework reveals distinct operational regimes and transition boundaries between retrieval-based and preprocessing-based strategies. Results show that deployment-aware optimization reduces effective token usage by approximately 25% at comparable performance ($F1 \approx 0.78$), while amortized memory compression achieves over 50% lower token cost relative to full-context prompting in higher-performance settings. Overall, the proposed framework provides a principled and practical foundation for evaluating and deploying scalable, efficient, and sustainable LLM systems.

Hanyu Cai, Binqi Shen, Lier Jin et al.

Prompt engineering has emerged as a critical factor influencing large language model (LLM) performance, yet the impact of pragmatic elements such as linguistic tone and politeness remains underexplored, particularly across different model families. In this work, we propose a systematic evaluation framework to examine how interaction tone affects model accuracy and apply it to three recently released and widely available LLMs: GPT-4o mini (OpenAI), Gemini 2.0 Flash (Google DeepMind), and Llama 4 Scout (Meta). Using the MMMLU benchmark, we evaluate model performance under Very Friendly, Neutral, and Very Rude prompt variants across six tasks spanning STEM and Humanities domains, and analyze pairwise accuracy differences with statistical significance testing. Our results show that tone sensitivity is both model-dependent and domain-specific. Neutral or Very Friendly prompts generally yield higher accuracy than Very Rude prompts, but statistically significant effects appear only in a subset of Humanities tasks, where rude tone reduces accuracy for GPT and Llama, while Gemini remains comparatively tone-insensitive. When performance is aggregated across tasks within each domain, tone effects diminish and largely lose statistical significance. Compared with earlier researches, these findings suggest that dataset scale and coverage materially influence the detection of tone effects. Overall, our study indicates that while interaction tone can matter in specific interpretive settings, modern LLMs are broadly robust to tonal variation in typical mixed-domain use, providing practical guidance for prompt design and model selection in real-world deployments.

Wanting Xu, Lan Hu, Manolis C. Tsakiris et al.

Over the past decade, the Gröbner basis theory and automatic solver generation have lead to a large number of solutions to geometric vision problems. In practically all cases, the derived solvers apply a fixed elimination template to calculate the Gröbner basis and thereby identify the zero-dimensional variety of the original polynomial constraints. However, it is clear that different variable or monomial orderings lead to different elimination templates, and we show that they may present a large variability in accuracy for a certain instance of a problem. The present paper has two contributions. We first show that for a common class of problems in geometric vision, variable reordering simply translates into a permutation of the columns of the initial coefficient matrix, and that -- as a result -- one and the same elimination template can be reused in different ways, each one leading to potentially different accuracy. We then prove that the original set of coefficients may contain sufficient information to train a classifier for online selection of a good solver, most notably at the cost of only a small computational overhead. We demonstrate wide applicability at the hand of generic dense polynomial problem solvers, as well as a concrete solver from geometric vision.

Lan Hu, Haomin Shi, Laurent Kneip

The present work proposes a solution to the challenging problem of registering two partial point sets of the same object with very limited overlap. We leverage the fact that most objects found in man-made environments contain a plane of symmetry. By reflecting the points of each set with respect to the plane of symmetry, we can largely increase the overlap between the sets and therefore boost the registration process. However, prior knowledge about the plane of symmetry is generally unavailable or at least very hard to find, especially with limited partial views, and finding this plane could strongly benefit from a prior alignment of the partial point sets. We solve this chicken-and-egg problem by jointly optimizing the relative pose and symmetry plane parameters, and notably do so under global optimality by employing the branch-and-bound (BnB) paradigm. Our results demonstrate a great improvement over the current state-of-the-art in globally optimal point set registration for common objects. We furthermore show an interesting application of our method to dense 3D reconstruction of scenes with repetitive objects.

Lan Hu, Wanting Xu, Kun Huang et al.

In an effort to increase the capabilities of SLAM systems and produce object-level representations, the community increasingly investigates the imposition of higher-level priors into the estimation process. One such example is given by employing object detectors to load and register full CAD models. Our work extends this idea to environments with unknown objects and imposes object priors by employing modern class-specific neural networks to generate complete model geometry proposals. The difficulty of using such predictions in a real SLAM scenario is that the prediction performance depends on the view-point and measurement quality, with even small changes of the input data sometimes leading to a large variability in the network output. We propose a discrete selection strategy that finds the best among multiple proposals from different registered views by re-enforcing the agreement with the online depth measurements. The result is an effective object-level RGBD SLAM system that produces compact, high-fidelity, and dense 3D maps with semantic annotations. It outperforms traditional fusion strategies in terms of map completeness and resilience against degrading measurement quality.

Lan Hu, Yuchen Cao, Peng Wu et al.

Most problems involving simultaneous localization and mapping can nowadays be solved using one of two fundamentally different approaches. The traditional approach is given by a least-squares objective, which minimizes many local photometric or geometric residuals over explicitly parametrized structure and camera parameters. Unmodeled effects violating the lambertian surface assumption or geometric invariances of individual residuals are encountered through statistical averaging or the addition of robust kernels and smoothness terms. Aiming at more accurate measurement models and the inclusion of higher-order shape priors, the community more recently shifted its attention to deep end-to-end models for solving geometric localization and mapping problems. However, at test-time, these feed-forward models ignore the more traditional geometric or photometric consistency terms, thus leading to a low ability to recover fine details and potentially complete failure in corner case scenarios. With an application to dense object modeling from RGBD images, our work aims at taking the best of both worlds by embedding modern higher-order object shape priors into classical iterative residual minimization objectives. We demonstrate a general ability to improve mapping accuracy with respect to each modality alone, and present a successful application to real data.