Align$^2$LLaVA: Cascaded Human and Large Language Model Preference Alignment for Multi-modal Instruction Curation

Recent advances in Multi-modal Large Language Models (MLLMs), such as LLaVA-series models, are driven by massive machine-generated instruction-following data tuning. Such automatic instruction collection pipelines, however, inadvertently introduce significant variability in data quality. This paper introduces a novel instruction curation algorithm, derived from two unique perspectives, human and LLM preference alignment, to compress this vast corpus of machine-generated multimodal instructions to a compact and high-quality form: (i) For human preference alignment, we have collected a machine-generated multimodal instruction dataset and established a comprehensive set of both subjective and objective criteria to guide the data quality assessment critically from human experts. By doing so, a reward model was trained on the annotated dataset to internalize the nuanced human understanding of instruction alignment. (ii) For LLM preference alignment, given the instruction selected by the reward model, we propose leveraging the inner LLM used in MLLM to align the writing style of visual instructions with that of the inner LLM itself, resulting in LLM-aligned instruction improvement. Extensive experiments demonstrate that we can maintain or even improve model performance by compressing synthetic multimodal instructions by up to 90%. Impressively, by aggressively reducing the total training sample size from 158k to 14k (9$\times$ smaller), our model consistently outperforms its full-size dataset counterpart across various MLLM benchmarks. Our project is available at https://github.com/DCDmllm/Align2LLaVA.

EMOVA: Empowering Language Models to See, Hear and Speak with Vivid Emotions

GPT-4o, an omni-modal model that enables vocal conversations with diverse emotions and tones, marks a milestone for omni-modal foundation models. However, empowering Large Language Models to perceive and generate images, texts, and speeches end-to-end with publicly available data remains challenging in the open-source community. Existing vision-language models rely on external tools for the speech processing, while speech-language models still suffer from limited or even without vision-understanding abilities. To address this gap, we propose EMOVA (EMotionally Omni-present Voice Assistant), to enable Large Language Models with end-to-end speech capabilities while maintaining the leading vision-language performance. With a semantic-acoustic disentangled speech tokenizer, we notice surprisingly that omni-modal alignment can further enhance vision-language and speech abilities compared with the corresponding bi-modal aligned counterparts. Moreover, a lightweight style module is proposed for flexible speech style controls (e.g., emotions and pitches). For the first time, EMOVA achieves state-of-the-art performance on both the vision-language and speech benchmarks, and meanwhile, supporting omni-modal spoken dialogue with vivid emotions.

Collaboratively Learning Federated Models from Noisy Decentralized Data

Federated learning (FL) has emerged as a prominent method for collaboratively training machine learning models using local data from edge devices, all while keeping data decentralized. However, accounting for the quality of data contributed by local clients remains a critical challenge in FL, as local data are often susceptible to corruption by various forms of noise and perturbations, which compromise the aggregation process and lead to a subpar global model. In this work, we focus on addressing the problem of noisy data in the input space, an under-explored area compared to the label noise. We propose a comprehensive assessment of client input in the gradient space, inspired by the distinct disparity observed between the density of gradient norm distributions of models trained on noisy and clean input data. Based on this observation, we introduce a straightforward yet effective approach to identify clients with low-quality data at the initial stage of FL. Furthermore, we propose a noise-aware FL aggregation method, namely Federated Noise-Sifting (FedNS), which can be used as a plug-in approach in conjunction with widely used FL strategies. Our extensive evaluation on diverse benchmark datasets under different federated settings demonstrates the efficacy of FedNS. Our method effortlessly integrates with existing FL strategies, enhancing the global model's performance by up to 13.68% in IID and 15.85% in non-IID settings when learning from noisy decentralized data.

LLaVA-MoD: Making LLaVA Tiny via MoE Knowledge Distillation

We introduce LLaVA-MoD, a novel framework designed to enable the efficient training of small-scale Multimodal Language Models (s-MLLM) by distilling knowledge from large-scale MLLM (l-MLLM). Our approach tackles two fundamental challenges in MLLM distillation. First, we optimize the network structure of s-MLLM by integrating a sparse Mixture of Experts (MoE) architecture into the language model, striking a balance between computational efficiency and model expressiveness. Second, we propose a progressive knowledge transfer strategy to ensure comprehensive knowledge migration. This strategy begins with mimic distillation, where we minimize the Kullback-Leibler (KL) divergence between output distributions to enable the student model to emulate the teacher network's understanding. Following this, we introduce preference distillation via Direct Preference Optimization (DPO), where the key lies in treating l-MLLM as the reference model. During this phase, the s-MLLM's ability to discriminate between superior and inferior examples is significantly enhanced beyond l-MLLM, leading to a better student that surpasses its teacher, particularly in hallucination benchmarks. Extensive experiments demonstrate that LLaVA-MoD outperforms existing models across various multimodal benchmarks while maintaining a minimal number of activated parameters and low computational costs. Remarkably, LLaVA-MoD, with only 2B activated parameters, surpasses Qwen-VL-Chat-7B by an average of 8.8% across benchmarks, using merely 0.3% of the training data and 23% trainable parameters. These results underscore LLaVA-MoD's ability to effectively distill comprehensive knowledge from its teacher model, paving the way for the development of more efficient MLLMs. The code will be available on: https://github.com/shufangxun/LLaVA-MoD.

NAS-Cap: Deep-Learning Driven 3-D Capacitance Extraction with Neural Architecture Search and Data Augmentation

More accurate capacitance extraction is demanded for designing integrated circuits under advanced process technology. The pattern matching approach and the field solver for capacitance extraction have the drawbacks of inaccuracy and large computational cost, respectively. Recent work \cite{yang2023cnn} proposes a grid-based data representation and a convolutional neural network (CNN) based capacitance models (called CNN-Cap), which opens the third way for 3-D capacitance extraction to get accurate results with much less time cost than field solver. In this work, the techniques of neural architecture search (NAS) and data augmentation are proposed to train better CNN models for 3-D capacitance extraction. Experimental results on datasets from different designs show that the obtained NAS-Cap models achieve remarkably higher accuracy than CNN-Cap, while consuming less runtime for inference and space for model storage. Meanwhile, the transferability of the NAS is validated, as the once searched architecture brought similar error reduction on coupling/total capacitance for the test cases from different design and/or process technology.

TeamLoRA: Boosting Low-Rank Adaptation with Expert Collaboration and Competition

While Parameter-Efficient Fine-Tuning (PEFT) methods like LoRA have effectively addressed GPU memory constraints during fine-tuning, their performance often falls short, especially in multidimensional task scenarios. To address this issue, one straightforward solution is to introduce task-specific LoRA modules as domain experts, leveraging the modeling of multiple experts' capabilities and thus enhancing the general capability of multi-task learning. Despite promising, these additional components often add complexity to the training and inference process, contravening the efficient characterization of PEFT designed for. Considering this, we introduce an innovative PEFT method, TeamLoRA, consisting of a collaboration and competition module for experts, and thus achieving the right balance of effectiveness and efficiency: (i) For collaboration, a novel knowledge-sharing and -organizing mechanism is devised to appropriately reduce the scale of matrix operations, thereby boosting the training and inference speed. (ii) For competition, we propose leveraging a game-theoretic interaction mechanism for experts, encouraging experts to transfer their domain-specific knowledge while facing diverse downstream tasks, and thus enhancing the performance. By doing so, TeamLoRA elegantly connects the experts as a "Team" with internal collaboration and competition, enabling a faster and more accurate PEFT paradigm for multi-task learning. To validate the superiority of TeamLoRA, we curate a comprehensive multi-task evaluation(CME) benchmark to thoroughly assess the capability of multi-task learning. Experiments conducted on our CME and other benchmarks indicate the effectiveness and efficiency of TeamLoRA. Our project is available at https://github.com/Lin-Tianwei/TeamLoRA.

MARS: Mixture of Auto-Regressive Models for Fine-grained Text-to-image Synthesis

Auto-regressive models have made significant progress in the realm of language generation, yet they do not perform on par with diffusion models in the domain of image synthesis. In this work, we introduce MARS, a novel framework for T2I generation that incorporates a specially designed Semantic Vision-Language Integration Expert (SemVIE). This innovative component integrates pre-trained LLMs by independently processing linguistic and visual information, freezing the textual component while fine-tuning the visual component. This methodology preserves the NLP capabilities of LLMs while imbuing them with exceptional visual understanding. Building upon the powerful base of the pre-trained Qwen-7B, MARS stands out with its bilingual generative capabilities corresponding to both English and Chinese language prompts and the capacity for joint image and text generation. The flexibility of this framework lends itself to migration towards any-to-any task adaptability. Furthermore, MARS employs a multi-stage training strategy that first establishes robust image-text alignment through complementary bidirectional tasks and subsequently concentrates on refining the T2I generation process, significantly augmenting text-image synchrony and the granularity of image details. Notably, MARS requires only 9% of the GPU days needed by SD1.5, yet it achieves remarkable results across a variety of benchmarks, illustrating the training efficiency and the potential for swift deployment in various applications.

EAGER: Two-Stream Generative Recommender with Behavior-Semantic Collaboration

Generative retrieval has recently emerged as a promising approach to sequential recommendation, framing candidate item retrieval as an autoregressive sequence generation problem. However, existing generative methods typically focus solely on either behavioral or semantic aspects of item information, neglecting their complementary nature and thus resulting in limited effectiveness. To address this limitation, we introduce EAGER, a novel generative recommendation framework that seamlessly integrates both behavioral and semantic information. Specifically, we identify three key challenges in combining these two types of information: a unified generative architecture capable of handling two feature types, ensuring sufficient and independent learning for each type, and fostering subtle interactions that enhance collaborative information utilization. To achieve these goals, we propose (1) a two-stream generation architecture leveraging a shared encoder and two separate decoders to decode behavior tokens and semantic tokens with a confidence-based ranking strategy; (2) a global contrastive task with summary tokens to achieve discriminative decoding for each type of information; and (3) a semantic-guided transfer task designed to implicitly promote cross-interactions through reconstruction and estimation objectives. We validate the effectiveness of EAGER on four public benchmarks, demonstrating its superior performance compared to existing methods.

T2S-GPT: Dynamic Vector Quantization for Autoregressive Sign Language Production from Text

In this work, we propose a two-stage sign language production (SLP) paradigm that first encodes sign language sequences into discrete codes and then autoregressively generates sign language from text based on the learned codebook. However, existing vector quantization (VQ) methods are fixed-length encodings, overlooking the uneven information density in sign language, which leads to under-encoding of important regions and over-encoding of unimportant regions. To address this issue, we propose a novel dynamic vector quantization (DVA-VAE) model that can dynamically adjust the encoding length based on the information density in sign language to achieve accurate and compact encoding. Then, a GPT-like model learns to generate code sequences and their corresponding durations from spoken language text. Extensive experiments conducted on the PHOENIX14T dataset demonstrate the effectiveness of our proposed method. To promote sign language research, we propose a new large German sign language dataset, PHOENIX-News, which contains 486 hours of sign language videos, audio, and transcription texts.Experimental analysis on PHOENIX-News shows that the performance of our model can be further improved by increasing the size of the training data. Our project homepage is https://t2sgpt-demo.yinaoxiong.cn.

CFMW: Cross-modality Fusion Mamba for Multispectral Object Detection under Adverse Weather Conditions

Cross-modality images that integrate visible-infrared spectra cues can provide richer complementary information for object detection. Despite this, existing visible-infrared object detection methods severely degrade in severe weather conditions. This failure stems from the pronounced sensitivity of visible images to environmental perturbations, such as rain, haze, and snow, which frequently cause false negatives and false positives in detection. To address this issue, we introduce a novel and challenging task, termed visible-infrared object detection under adverse weather conditions. To foster this task, we have constructed a new Severe Weather Visible-Infrared Dataset (SWVID) with diverse severe weather scenes. Furthermore, we introduce the Cross-modality Fusion Mamba with Weather-removal (CFMW) to augment detection accuracy in adverse weather conditions. Thanks to the proposed Weather Removal Diffusion Model (WRDM) and Cross-modality Fusion Mamba (CFM) modules, CFMW is able to mine more essential information of pedestrian features in cross-modality fusion, thus could transfer to other rarer scenarios with high efficiency and has adequate availability on those platforms with low computing power. To the best of our knowledge, this is the first study that targeted improvement and integrated both Diffusion and Mamba modules in cross-modality object detection, successfully expanding the practical application of this type of model with its higher accuracy and more advanced architecture. Extensive experiments on both well-recognized and self-created datasets conclusively demonstrate that our CFMW achieves state-of-the-art detection performance, surpassing existing benchmarks. The dataset and source code will be made publicly available at https://github.com/lhy-zjut/CFMW.