Haoyu Song

Weizhi Wang, Li Dong, Hao Cheng et al. · microsoft-research

Human language is grounded on multimodal knowledge including visual knowledge like colors, sizes, and shapes. However, current large-scale pre-trained language models rely on text-only self-supervised training with massive text data, which precludes them from utilizing relevant visual information when necessary. To address this, we propose a novel pre-training framework, named VaLM, to Visually-augment text tokens with retrieved relevant images for Language Modeling. Specifically, VaLM builds on a novel latent text-image alignment method via an image retrieval module to fetch corresponding images given a textual context. With the visually-augmented context, VaLM uses a visual knowledge fusion layer to enable multimodal grounded language modeling by attending to both text context and visual knowledge in images. We evaluate VaLM on various visual knowledge-intensive commonsense reasoning tasks, which require visual information to excel. The experimental results illustrate that VaLM outperforms all strong language-only and vision-language baselines with substantial gains in reasoning object commonsense including color, size, and shape. Our code is available at https://github.com/Victorwz/VaLM.

Haoyu Song, Li Dong, Wei-Nan Zhang et al. · microsoft-research

CLIP has shown a remarkable zero-shot capability on a wide range of vision tasks. Previously, CLIP is only regarded as a powerful visual encoder. However, after being pre-trained by language supervision from a large amount of image-caption pairs, CLIP itself should also have acquired some few-shot abilities for vision-language tasks. In this work, we empirically show that CLIP can be a strong vision-language few-shot learner by leveraging the power of language. We first evaluate CLIP's zero-shot performance on a typical visual question answering task and demonstrate a zero-shot cross-modality transfer capability of CLIP on the visual entailment task. Then we propose a parameter-efficient fine-tuning strategy to boost the few-shot performance on the vqa task. We achieve competitive zero/few-shot results on the visual question answering and visual entailment tasks without introducing any additional pre-training procedure.

Yaru Hao, Haoyu Song, Li Dong et al. · microsoft-research

Foundation models have received much attention due to their effectiveness across a broad range of downstream applications. Though there is a big convergence in terms of architecture, most pretrained models are typically still developed for specific tasks or modalities. In this work, we propose to use language models as a general-purpose interface to various foundation models. A collection of pretrained encoders perceive diverse modalities (such as vision, and language), and they dock with a language model that plays the role of a universal task layer. We propose a semi-causal language modeling objective to jointly pretrain the interface and the modular encoders. We subsume the advantages and capabilities from both causal and non-causal modeling, thereby combining the best of two worlds. Specifically, the proposed method not only inherits the capabilities of in-context learning and open-ended generation from causal language modeling, but also is conducive to finetuning because of the bidirectional encoders. More importantly, our approach seamlessly unlocks the combinations of the above capabilities, e.g., enabling in-context learning or instruction following with finetuned encoders. Experimental results across various language-only and vision-language benchmarks show that our model outperforms or is competitive with specialized models on finetuning, zero-shot generalization, and few-shot learning.

Xiao Wang, Haiyang Wang, Shiao Wang et al.

Existing pedestrian attribute recognition methods are generally developed based on RGB frame cameras. However, these approaches are constrained by the limitations of RGB cameras, such as sensitivity to lighting conditions and motion blur, which hinder their performance. Furthermore, current attribute recognition primarily focuses on analyzing pedestrians' external appearance and clothing, lacking an exploration of emotional dimensions. In this paper, we revisit these issues and propose a novel multi-modal RGB-Event attribute recognition task by drawing inspiration from the advantages of event cameras in low-light, high-speed, and low-power consumption. Specifically, we introduce the first large-scale multi-modal pedestrian attribute recognition dataset, termed EventPAR, comprising 100K paired RGB-Event samples that cover 50 attributes related to both appearance and six human emotions, diverse scenes, and various seasons. By retraining and evaluating mainstream PAR models on this dataset, we establish a comprehensive benchmark and provide a solid foundation for future research in terms of data and algorithmic baselines. In addition, we propose a novel RWKV-based multi-modal pedestrian attribute recognition framework, featuring an RWKV visual encoder and an asymmetric RWKV fusion module. Extensive experiments are conducted on our proposed dataset as well as two simulated datasets (MARS-Attribute and DukeMTMC-VID-Attribute), achieving state-of-the-art results. The source code and dataset will be released on https://github.com/Event-AHU/OpenPAR

Lan Chen, Haoxiang Yang, Pengpeng Shao et al.

Pattern recognition leveraging both RGB and Event cameras can significantly enhance performance by deploying deep neural networks that utilize a fine-tuning strategy. Inspired by the successful application of large models, the introduction of such large models can also be considered to further enhance the performance of multi-modal tasks. However, fully fine-tuning these models leads to inefficiency and lightweight fine-tuning methods such as LoRA and Adapter have been proposed to achieve a better balance between efficiency and performance. To our knowledge, there is currently no work that has conducted parameter-efficient fine-tuning (PEFT) for RGB-Event recognition based on pre-trained foundation models. To address this issue, this paper proposes a novel PEFT strategy to adapt the pre-trained foundation vision models for the RGB-Event-based classification. Specifically, given the RGB frames and event streams, we extract the RGB and event features based on the vision foundation model ViT with a modality-specific LoRA tuning strategy. The frame difference of the dual modalities is also considered to capture the motion cues via the frame difference backbone network. These features are concatenated and fed into high-level Transformer layers for efficient multi-modal feature learning via modality-shared LoRA tuning. Finally, we concatenate these features and feed them into a classification head to achieve efficient fine-tuning. The source code and pre-trained models will be released on \url{https://github.com/Event-AHU/VELoRA}.

Crystal Min Hui Poon, Pai Chet Ng, Xiaoxiao Miao et al.

Instruction-guided text-to-speech (TTS) research has reached a maturity level where excellent speech generation quality is possible on demand, yet two coupled biases persist: accent bias, where models default to dominant phonetic patterns, and linguistic bias, where dialect-specific lexical and cultural cues are ignored. These biases are interdependent, as authentic accent generation requires both accent fidelity and localized text. We present Contextual Linguistic Adaptation and Retrieval for Inclusive TTS sYnthesis (CLARITY), a backbone-agnostic framework that addresses these biases through dual-signal optimization: (i) contextual linguistic adaptation that localizes input text to the target dialect, and (ii) retrieval-augmented accent prompting (RAAP) that supplies accent-consistent speech prompts. Across twelve English accents, CLARITY improves accent accuracy and fairness while maintaining strong perceptual quality.