Robust Few-Shot Vision-Language Model Adaptation

Pretrained VLMs achieve strong performance on downstream tasks when adapted with just a few labeled examples. As the adapted models inevitably encounter out-of-distribution (OOD) test data that deviates from the in-distribution (ID) task-specific training data, enhancing OOD generalization in few-shot adaptation is critically important. We study robust few-shot VLM adaptation, aiming to increase both ID and OOD accuracy. By comparing different adaptation methods (e.g., prompt tuning, linear probing, contrastive finetuning, and full finetuning), we uncover three key findings: (1) finetuning with proper hyperparameters significantly outperforms the popular VLM adaptation methods prompt tuning and linear probing; (2) visual encoder-only finetuning achieves better efficiency and accuracy than contrastively finetuning both visual and textual encoders; (3) finetuning the top layers of the visual encoder provides the best balance between ID and OOD accuracy. Building on these findings, we propose partial finetuning of the visual encoder empowered with two simple augmentation techniques: (1) retrieval augmentation which retrieves task-relevant data from the VLM's pretraining dataset to enhance adaptation, and (2) adversarial perturbation which promotes robustness during finetuning. Results show that the former/latter boosts OOD/ID accuracy while slightly sacrificing the ID/OOD accuracy. Yet, perhaps understandably, naively combining the two does not maintain their best OOD/ID accuracy. We address this dilemma with the developed SRAPF, Stage-wise Retrieval Augmentation-based Adversarial Partial Finetuning. SRAPF consists of two stages: (1) partial finetuning the visual encoder using both ID and retrieved data, and (2) adversarial partial finetuning with few-shot ID data. Extensive experiments demonstrate that SRAPF achieves the state-of-the-art ID and OOD accuracy on the ImageNet OOD benchmarks.

CAST: Contrastive Adaptation and Distillation for Semi-Supervised Instance Segmentation

Instance segmentation demands costly per-pixel annotations and large models. We introduce CAST, a semi-supervised knowledge distillation (SSKD) framework that compresses pretrained vision foundation models (VFM) into compact experts using limited labeled and abundant unlabeled data. CAST unfolds in three stages: (1) domain adaptation of the VFM teacher(s) via self-training with contrastive pixel calibration, (2) distillation into a compact student via a unified multi-objective loss that couples standard supervision and pseudo-labels with our instance-aware pixel-wise contrastive term, and (3) fine-tuning on labeled data to remove residual pseudo-label bias. Central to CAST is an \emph{instance-aware pixel-wise contrastive loss} that fuses mask and class scores to mine informative negatives and enforce clear inter-instance margins. By maintaining this contrastive signal across both adaptation and distillation, we align teacher and student embeddings and fully leverage unlabeled images. On Cityscapes and ADE20K, our ~11X smaller student surpasses its adapted VFM teacher(s) by +3.4 AP (33.9 vs. 30.5) and +1.5 AP (16.7 vs. 15.2) and outperforms state-of-the-art semi-supervised approaches.

Confidence in Large Language Model Evaluation: A Bayesian Approach to Limited-Sample Challenges

Large language models (LLMs) exhibit probabilistic output characteristics, yet conventional evaluation frameworks rely on deterministic scalar metrics. This study introduces a Bayesian approach for LLM capability assessment that integrates prior knowledge through probabilistic inference, addressing limitations under limited-sample regimes. By treating model capabilities as latent variables and leveraging a curated query set to induce discriminative responses, we formalize model ranking as a Bayesian hypothesis testing problem over mutually exclusive capability intervals. Experimental evaluations with GPT-series models demonstrate that the proposed method achieves superior discrimination compared to conventional evaluation methods. Results indicate that even with reduced sample sizes, the approach maintains statistical robustness while providing actionable insights, such as probabilistic statements about a model's likelihood of surpassing specific baselines. This work advances LLM evaluation methodologies by bridging Bayesian inference with practical constraints in real-world deployment scenarios.

Weakly Supervised Spatial Implicit Neural Representation Learning for 3D MRI-Ultrasound Deformable Image Registration in HDR Prostate Brachytherapy

Purpose: Accurate 3D MRI-ultrasound (US) deformable registration is critical for real-time guidance in high-dose-rate (HDR) prostate brachytherapy. We present a weakly supervised spatial implicit neural representation (SINR) method to address modality differences and pelvic anatomy challenges. Methods: The framework uses sparse surface supervision from MRI/US segmentations instead of dense intensity matching. SINR models deformations as continuous spatial functions, with patient-specific surface priors guiding a stationary velocity field for biologically plausible deformations. Validation included 20 public Prostate-MRI-US-Biopsy cases and 10 institutional HDR cases, evaluated via Dice similarity coefficient (DSC), mean surface distance (MSD), and 95% Hausdorff distance (HD95). Results: The proposed method achieved robust registration. For the public dataset, prostate DSC was $0.93 \pm 0.05$, MSD $0.87 \pm 0.10$ mm, and HD95 $1.58 \pm 0.37$ mm. For the institutional dataset, prostate CTV achieved DSC $0.88 \pm 0.09$, MSD $1.21 \pm 0.38$ mm, and HD95 $2.09 \pm 1.48$ mm. Bladder and rectum performance was lower due to ultrasound's limited field of view. Visual assessments confirmed accurate alignment with minimal discrepancies. Conclusion: This study introduces a novel weakly supervised SINR-based approach for 3D MRI-US deformable registration. By leveraging sparse surface supervision and spatial priors, it achieves accurate, robust, and computationally efficient registration, enhancing real-time image guidance in HDR prostate brachytherapy and improving treatment precision.

Hunyuan3D 2.0: Scaling Diffusion Models for High Resolution Textured 3D Assets Generation

We present Hunyuan3D 2.0, an advanced large-scale 3D synthesis system for generating high-resolution textured 3D assets. This system includes two foundation components: a large-scale shape generation model -- Hunyuan3D-DiT, and a large-scale texture synthesis model -- Hunyuan3D-Paint. The shape generative model, built on a scalable flow-based diffusion transformer, aims to create geometry that properly aligns with a given condition image, laying a solid foundation for downstream applications. The texture synthesis model, benefiting from strong geometric and diffusion priors, produces high-resolution and vibrant texture maps for either generated or hand-crafted meshes. Furthermore, we build Hunyuan3D-Studio -- a versatile, user-friendly production platform that simplifies the re-creation process of 3D assets. It allows both professional and amateur users to manipulate or even animate their meshes efficiently. We systematically evaluate our models, showing that Hunyuan3D 2.0 outperforms previous state-of-the-art models, including the open-source models and closed-source models in geometry details, condition alignment, texture quality, and etc. Hunyuan3D 2.0 is publicly released in order to fill the gaps in the open-source 3D community for large-scale foundation generative models. The code and pre-trained weights of our models are available at: https://github.com/Tencent/Hunyuan3D-2

A Real-Time System for Scheduling and Managing UAV Delivery in Urban

As urban logistics demand continues to grow, UAV delivery has become a key solution to improve delivery efficiency, reduce traffic congestion, and lower logistics costs. However, to fully leverage the potential of UAV delivery networks, efficient swarm scheduling and management are crucial. In this paper, we propose a real-time scheduling and management system based on the ``Airport-Unloading Station" model, aiming to bridge the gap between high-level scheduling algorithms and low-level execution systems. This system, acting as middleware, accurately translates the requirements from the scheduling layer into specific execution instructions, ensuring that the scheduling algorithms perform effectively in real-world environments. Additionally, we implement three collaborative scheduling schemes involving autonomous ground vehicles (AGVs), unmanned aerial vehicles (UAVs), and ground staff to further optimize overall delivery efficiency. Through extensive experiments, this study demonstrates the rationality and feasibility of the proposed management system, providing practical solution for the commercial application of UAVs delivery in urban. Code: https://github.com/chengji253/UAVDeliverySystem

UAL-Bench: The First Comprehensive Unusual Activity Localization Benchmark

Localizing unusual activities, such as human errors or surveillance incidents, in videos holds practical significance. However, current video understanding models struggle with localizing these unusual events likely because of their insufficient representation in models' pretraining datasets. To explore foundation models' capability in localizing unusual activity, we introduce UAL-Bench, a comprehensive benchmark for unusual activity localization, featuring three video datasets: UAG-OOPS, UAG-SSBD, UAG-FunQA, and an instruction-tune dataset: OOPS-UAG-Instruct, to improve model capabilities. UAL-Bench evaluates three approaches: Video-Language Models (Vid-LLMs), instruction-tuned Vid-LLMs, and a novel integration of Vision-Language Models and Large Language Models (VLM-LLM). Our results show the VLM-LLM approach excels in localizing short-span unusual events and predicting their onset (start time) more accurately than Vid-LLMs. We also propose a new metric, R@1, TD <= p, to address limitations in existing evaluation methods. Our findings highlight the challenges posed by long-duration videos, particularly in autism diagnosis scenarios, and the need for further advancements in localization techniques. Our work not only provides a benchmark for unusual activity localization but also outlines the key challenges for existing foundation models, suggesting future research directions on this important task.

Few-Shot Recognition via Stage-Wise Augmented Finetuning

Few-shot recognition aims to train a classification model with only a few labeled examples of pre-defined concepts, where annotation can be costly in a downstream task. In another related research area, zero-shot recognition, which assumes no access to any downstream-task data, has been greatly advanced by using pretrained Vision-Language Models (VLMs). In this area, retrieval-augmented learning (RAL) effectively boosts zero-shot accuracy by retrieving and learning from external data relevant to downstream concepts. Motivated by these advancements, our work explores RAL for few-shot recognition. While seemingly straightforward despite being under-explored in the literature (till now!), we present novel challenges and opportunities for applying RAL for few-shot recognition. First, perhaps surprisingly, simply finetuning the VLM on a large amount of retrieved data barely surpasses state-of-the-art zero-shot methods due to the imbalanced distribution of retrieved data and its domain gaps compared to few-shot annotated data. Second, finetuning a VLM on few-shot examples alone significantly outperforms prior methods, and finetuning on the mix of retrieved and few-shot data yields even better results. Third, to mitigate the imbalanced distribution and domain gap issue, we propose Stage-Wise Augmented fineTuning (SWAT) method, which involves end-to-end finetuning on mixed data for the first stage and retraining the classifier solely on the few-shot data in the second stage. Extensive experiments show that SWAT achieves the best performance on standard benchmark datasets, resoundingly outperforming prior works by ~10% in accuracy. Code is available at https://github.com/tian1327/SWAT.

Learning using granularity statistical invariants for classification

Learning using statistical invariants (LUSI) is a new learning paradigm, which adopts weak convergence mechanism, and can be applied to a wider range of classification problems. However, the computation cost of invariant matrices in LUSI is high for large-scale datasets during training. To settle this issue, this paper introduces a granularity statistical invariant for LUSI, and develops a new learning paradigm called learning using granularity statistical invariants (LUGSI). LUGSI employs both strong and weak convergence mechanisms, taking a perspective of minimizing expected risk. As far as we know, it is the first time to construct granularity statistical invariants. Compared to LUSI, the introduction of this new statistical invariant brings two advantages. Firstly, it enhances the structural information of the data. Secondly, LUGSI transforms a large invariant matrix into a smaller one by maximizing the distance between classes, achieving feasibility for large-scale datasets classification problems and significantly enhancing the training speed of model operations. Experimental results indicate that LUGSI not only exhibits improved generalization capabilities but also demonstrates faster training speed, particularly for large-scale datasets.

The Neglected Tails of Vision-Language Models

Vision-language models (VLMs) excel in zero-shot recognition but their performance varies greatly across different visual concepts. For example, although CLIP achieves impressive accuracy on ImageNet (60-80%), its performance drops below 10% for more than ten concepts like night snake, presumably due to their limited presence in the pretraining data. However, measuring the frequency of concepts in VLMs' large-scale datasets is challenging. We address this by using large language models (LLMs) to count the number of pretraining texts that contain synonyms of these concepts. Our analysis confirms that popular datasets, such as LAION, exhibit a long-tailed concept distribution, yielding biased performance in VLMs. We also find that downstream applications of VLMs, including visual chatbots (e.g., GPT-4V) and text-to-image models (e.g., Stable Diffusion), often fail to recognize or generate images of rare concepts identified by our method. To mitigate the imbalanced performance of zero-shot VLMs, we propose REtrieval-Augmented Learning (REAL). First, instead of prompting VLMs using the original class names, REAL uses their most frequent synonyms found in pretraining texts. This simple change already outperforms costly human-engineered and LLM-enriched prompts over nine benchmark datasets. Second, REAL trains a linear classifier on a small yet balanced set of pretraining data retrieved using concept synonyms. REAL surpasses the previous zero-shot SOTA, using 400x less storage and 10,000x less training time!