United States

Recent progress in Multimodal Large Language Models (MLLMs) often use large image tokens to compensate the visual shortcoming of MLLMs, which not only exhibits obvious redundancy but also greatly exacerbates the already high computation. Token pruning is an effective solution for speeding up MLLMs, but  when and how to drop tokens still remains a challenge.  In this paper, we propose a novel and training-free approach for the effective visual token pruning of MLLMs, termed FitPrune, which can quickly produce a complete pruning recipe for MLLMs according to a pre-defined budget.  Specifically, FitPrune considers token pruning as a statistical problem of MLLM and its objective is to find out an optimal pruning scheme that can minimize the divergence of the attention distributions before and after pruning. In practice, FitPrune can be quickly accomplished based on the attention statistics from a small batch of inference data, avoiding the expensive trials of MLLMs.
According to the pruning recipe, an MLLM can directly remove the redundant visual tokens of different examples during inference.  To validate FitPrune, we apply it to a set of recent MLLMs, including LLaVA-1.5, LLaVA-HR and LLaVA-NEXT, and conduct extensive experiments on a set of benchmarks. The experimental results show that our FitPrune can not only reduce the computational complexity to a large extent, while retaining high performance, e.g., -54.9\% FLOPs for LLaVA-NEXT with only 0.5\% accuracy drop. Notably, the pruning recipe can be obtained in about 5 minutes. Our code is available at 
https://anonymous.4open.science/r/AAAI25-B20E.

AAAI 2025

Fit and Prune: Fast and Training-free Visual Token Pruning for Multi-modal Large Language Models

language and vision

We are pleased to announce the Thirty-Ninth AAAI Conference on Artificial Intelligence (AAAI-25), which will be held in Philadelphia, Pennsylvania at the Pennsylvania Convention Center from February 25 to March 4, 2025.

The purpose of the AAAI conference series is to promote research in Artificial Intelligence (AI) and foster scientific exchange between researchers, practitioners, scientists, students, and engineers across the entirety of AI and its affiliated disciplines. AAAI-25 will feature technical paper presentations, special tracks, invited speakers, workshops, tutorials, poster sessions, senior member presentations, competitions, and exhibit programs, and a range of other activities to be announced.

### [Invited Speakers](https://aaai.org/conference/aaai/aaai-25/aaai-25-invited-speakers/)

Register [here](https://aaai.org/conference/aaai/aaai-25/registration/)

The purpose of the AAAI conference series is to promote research in Artificial Intelligence (AI) and foster scientific exchange between researchers, practitioners, scientists, students, and engineers across the entirety of AI and its affiliated disciplines. AAAI-25 will feature technical paper presentations, special tracks, invited speakers, workshops, tutorials, poster sessions, senior member presentations, competitions, and exhibit programs, and a range of other activities to be announced.



poster

Federated Learning (FL) offers a decentralized approach to model training, where data remains local and only model parameters are shared between the clients and the central server. 
Traditional methods, such as Federated Averaging (FedAvg), linearly average these parameters on the server, which often come from heterogeneous data distributions, potentially overlooking the complex, high-dimensional nature of the parameter space. 
This can result in poor generalization ability and reduced robustness. To address this issue, we propose a novel model aggregation framework, \algfont{FedGPA}. 
In this framework, the server deploys a diffusion model to consolidate uploaded parameters from all clients and efficiently generates personalized parameters for each client with global knowledge guidance using an inversion technique. 
Since the dependence of a client's model parameters on specific data distributions is encoded by a high-capacity diffusion model, \algfont{FedGPA} can effectively decouple the complexity of diverse data distributions from the complexity of individual clients.

pFedGPA: Diffusion-based Generative Parameter Aggregation for Personalized Federated Learning

Consistency regularization and pseudo-labeling have significantly advanced semi-supervised learning (SSL). Prior works have effectively employed Mixup for consistency regularization in SSL. However, our findings indicate that applying Mixup for consistency regularization may degrade SSL performance by compromising the purity of artificial labels. Moreover, most pseudo-labeling based methods utilize thresholding strategy to exclude low-confidence data, aiming to mitigate confirmation bias; however, this approach limits the utility of unlabeled samples. To address these challenges, we propose RegMixMatch, a novel framework that optimizes the use of Mixup with both high- and low-confidence samples in SSL. First, we introduce semi-supervised RegMixup, which effectively addresses reduced artificial labels purity by using both mixed samples and clean samples for training. Second, we develop a class-aware Mixup technique that integrates information from the top-2 predicted classes into low-confidence samples and their artificial labels, reducing the confirmation bias associated with these samples and enhancing their effective utilization. Experimental results demonstrate that RegMixMatch achieves state-of-the-art performance across various SSL benchmarks.

RegMixMatch: Optimizing Mixup Utilization in Semi-Supervised Learning

This paper studies the prediction task of tensor-on-tensor regression in which both covariates and responses are multi-dimensional arrays (a.k.a., tensors) across time with arbitrary tensor order and data dimension. Existing methods either focused on linear models without accounting for possibly nonlinear relationships between covariates and responses, or directly employed black-box deep learning algorithms that failed to utilize the inherent tensor structure. In this work, we propose a Factor Augmented Tensor-on-Tensor Neural Network (FATTNN) that integrates tensor factor models into deep neural networks. We begin with summarizing and extracting useful predictive information (represented by the ``factor tensor'') from the complex structured tensor covariates, and then proceed with the prediction task using the estimated factor tensor as input of a temporal convolutional neural network. The proposed methods effectively handle nonlinearity between complex data structures, and improve over traditional statistical models and conventional deep learning approaches in both prediction accuracy and computational cost. By leveraging tensor factor models, our proposed methods exploit the underlying latent factor structure to enhance the prediction, and in the meantime, drastically reduce the data dimensionality that speeds up the computation. The empirical performances of our proposed methods are demonstrated via simulation studies and real-world applications to three public datasets. Numerical results show that our proposed algorithms achieve substantial increases in prediction accuracy and significant reductions in computational time compared to benchmark methods.

Factor Augmented Tensor-on-Tensor Neural Networks

The performance of various tasks of natural language processing has greatly improved with the emergence of large language models. However, there is still much room for improvement in understanding certain specific linguistic phenomena, such as Chinese idioms, which are usually composed of four characters. Chinese idioms are difficult to understand due to semantic gaps between their literal and actual meanings. Researchers have proposed the Chinese idiom reading comprehension task to examine the ability of large language models to represent and understand Chinese idioms. The task requires choosing the correct Chinese idiom from a list of candidates to complete the sentence. The current research mainly focuses on text-based idiom comprehension. Nevertheless, there are many idiom application scenarios that combine images and text, and we believe that the corresponding images are beneficial for the model's understanding of the idioms. Therefore, to address the above problems, we first construct a large-scale Multimodal Chinese Idiom Reading Comprehension dataset (MChIRC), which contains a total of 44,433 image-text pairs covering 2,926 idioms. Then, we propose a Dual-Contrastive Idiom Graph Network (DCIGN), which employs a dual-contrastive learning module to align the text and image features corresponding to the same Chinese idiom at both coarse and fine levels, while utilizing a graph structure to capture the semantic relationships between idiom candidates. Finally, we use a cross-attention module to fuse multimodal features with graph features of candidate idioms to predict correct answers. The authoritativeness of MChIRC and the effectiveness of DCIGN are demonstrated through a variety of experiments, which provides a new benchmark for the multimodal Chinese idiom reading comprehension task.

McHirc: A Multimodal Benchmark for Chinese Idiom Reading Comprehension

Scene Graph Generation (SGG) research has suffered from two fundamental challenges: the long-tailed predicate distribution and semantic ambiguity between predicates. These challenges lead to a bias towards head predicates in SGG models, favoring dominant general predicates while overlooking fine-grained predicates. In this paper, we address the challenges of SGG by framing it as multi-label classification problem with partial annotation, where relevant labels of fine-grained predicates are missing. Under the new frame, we propose ReTrieval-Augmented scene graph Generation (ReTAG), which identifies potential instances to be multi-labeled and enriches the single-label with multi-labels that are semantically similar to the original label by retrieving relevant samples from our established memory bank. Based on augmented relations (i.e., discovered multi-labels), we apply multi-prototype learning to train our SGG model. Several comprehensive experiments have demonstrated that ReTAG outperforms state-of-the-art baselines by up to 3.6% on VG and 5.9% on GQA, particularly in terms of F@K, showing that ReTAG effectively alleviates the issue of biased prediction caused by the long-tailed distribution and semantic ambiguity of predicates.

RA-SGG: Retrieval-Augmented Scene Graph Generation Framework via Multi-Prototype Learning

Recent years have witnessed the remarkable success of Text-to-3D generation, particularly with the rise of mainstream conditional diffusion models (DMs). Though achieving substantial progress, existing methods still face a knotty $\textbf{\emph{``human preference''}}$ dilemma, that is the 3D contents generated by the models often deviate greatly from the desired effects (e.g., perspective, aesthetics, shading, appearance, etc.) due to the lack of attention to human preferences. To mitigate the limitation of data deficiency and enable human preference learning, we first elaborately curate the HP3D, a text-to-3D dataset with expert preference annotations which is initally captioned by the multimodal large model LLava and then refined by human expert. Based on such a brand-new HP3D, we further propose DreamAlign, a reward-free method that does not require designing any complex reward models whereas only by introducing a light-weight lora adapter and then designing a fresh direct 3D preference optimization (D-3DPO) algorithm for training. Moreover, in the stage of text-to-3D content generation we additional design Preference Contrastive Feedback training for score distillation sampling, which enables the generated 3D objects to align the human preferences (e.g., aesthetics, material, etc.).  Extensive experiments demonstrate that DreamAlign consistently achieves state-of-the-art performance both fine-grained generative effects and human preference alignment across the various benchmark evaluations.

DreamAlign: Dynamic Text-to-3D Optimization with Human Preference Alignment

While fine-tuned large language models (LLMs) excel in generating grammatically valid SQL in Text-to-SQL parsing, they often struggle to ensure semantic accuracy in queries, leading to user confusion and diminished system usability. To tackle this challenge, we introduce SQLFixAgent, a new consistency-enhanced multi-agent collaborative framework designed for detecting and repairing erroneous SQL. Our framework comprises a core agent, SQLRefiner, alongside two auxiliary agents: SQLReviewer and QueryCrafter. The SQLReviewer agent employs the rubber duck debugging method to identify potential semantic mismatches between SQL and user query. If the error is detected, the QueryCrafter agent generates multiple SQL as candidate repairs using a fine-tuned SQLTool. Subsequently, leveraging similar repair retrieval and failure memory reflection, the SQLRefiner agent selects the most fitting SQL statement from the candidates as the final repair. We evaluated our proposed framework on five Text-to-SQL benchmarks. The experimental results show that our method consistently enhances the performance of the baseline model, specifically achieving an execution accuracy improvement of over 3\% on the Bird benchmark. Our framework also has a higher token efficiency compared to other advanced methods, making it more competitive.

SQLFixAgent: Towards Semantic-Accurate Text-to-SQL Parsing via Consistency-Enhanced Multi-Agent Collaboration

The primary challenge of cross-domain few-shot segmentation (CD-FSS) is the domain disparity between the training and inference phases, which can exist in either the input data or the target classes. Previous models struggle to learn feature representations that generalize to various unknown domains from limited training domain samples. In contrast, the large-scale visual model SAM, pre-trained on tens of millions of images from various domains and classes, possesses excellent generalizability. In this work, we propose a SAM-aware graph prompt reasoning network (GPRN) that fully leverages SAM to guide CD-FSS feature representation learning and improve prediction accuracy. Specifically, we propose a SAM-aware prompt initialization module (SPI) to transform the masks generated by SAM into visual prompts enriched with high-level semantic information. Since SAM tends to divide an object into many sub-regions, this may lead to visual prompts representing the same semantic object having inconsistent or fragmented features. We further propose a graph prompt reasoning (GPR) module that constructs a graph among visual prompts to reason about their interrelationships and enable each visual prompt to aggregate information from similar prompts, thus achieving global semantic consistency. Subsequently, each visual prompt embeds its semantic information into the corresponding mask region to assist in feature representation learning. To refine the segmentation mask during testing, we also design a non-parameter adaptive point selection module (APS) to select representative point prompts from query predictions and feed them back to SAM to refine inaccurate segmentation results. Experiments on four standard CD-FSS datasets demonstrate that our method establishes new state-of-the-art results.

SAM-Aware Graph Prompt Reasoning Network for Cross-Domain Few-Shot Segmentation

Fine-tuning large language models (LLMs) requires significant memory, often exceeding the capacity of a single GPU. A common solution to this memory challenge is offloading compute and data from the GPU to the CPU. However, this approach is hampered by the limited bandwidth of commodity hardware, which constrains communication between the CPU and GPU, and by slower matrix multiplications on the CPU.

In this paper, we present an offloading framework, LSP-Offload, that enables near-native speed LLM fine-tuning on commodity hardware through learned sparse projectors. Our data-driven approach involves learning efficient sparse compressors that minimize communication with minimal precision loss. Additionally, we introduce a novel layer-wise communication schedule to maximize parallelism between communication and computation. As a result, our framework can fine-tune a 1.3 billion parameter model on a 4GB laptop GPU and a 6.7 billion parameter model on an NVIDIA RTX 4090 GPU with 24GB memory. Compared to state-of-the-art offloading frameworks, our approach reduces end-to-end fine-tuning time by 33.1%-62.5% when converging to the same accuracy.

Practical Offloading for Fine-Tuning LLM on Commodity GPU via Learned Subspace Projectors

Few-shot anomaly detection (FSAD) aims to detect unseen anomaly regions with the guidance of very few normal support images from the same class. Existing FSAD methods usually find anomalies by directly designing complex text prompts to align them with visual features under the prevailing large vision-language model paradigm. However, these methods, almost always, neglect intrinsic contextual information in visual features, e.g., the interaction relationships between different vision layers, which is an important clue for detecting anomalies comprehensively. To this end, we propose a kernel-aware graph prompt learning framework, termed as KAG-prompt, by reasoning the cross-layer relations among visual features for FSAD. Specifically, a kernel-aware hierarchical graph is built by taking the different layer features focusing on anomalous regions of different sizes as nodes, meanwhile, the relationships between arbitrary pairs of nodes stand for the edges of the graph. By message passing over this graph, KAG-prompt can capture cross-layer contextual information, thus leading to more accurate anomaly prediction. Moreover, to integrate the information of multiple important anomaly signals in the prediction map, we propose a novel image-level scoring method based on multi-level information fusion. Extensive experiments on MVTecAD and VisA datasets show that KAG-prompt achieves state-of-the-art FSAD results for image-level/pixel-level anomaly detection. Code is available at \url{https://anonymous.4open.science/r/KAG-prompt-3537}.

Downloads

Next from AAAI 2025

pFedGPA: Diffusion-based Generative Parameter Aggregation for Personalized Federated Learning

Similar lecture

What kind of visual tokens do we need? Training-free Visual Token Pruning for Multi-modal Large Language Models from the Perspective of Graph

.css-70qvj9{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;}Downloads

Next from AAAI 2025

pFedGPA: Diffusion-based Generative Parameter Aggregation for Personalized Federated Learning

Similar lecture

What kind of visual tokens do we need? Training-free Visual Token Pruning for Multi-modal Large Language Models from the Perspective of Graph

Downloads