Recent advancements have shown that the Mixture of Experts (MoE) approach significantly enhances the capacity of large language models (LLMs) and improves performance on downstream tasks. Building on these promising results, multi-modal large language models (MLLMs) have increasingly adopted MoE techniques. However, existing multi-modal MoE tuning methods typically face two key challenges: expert uniformity and router rigidity. Expert uniformity occurs because MoE experts are often initialized by simply replicating the FFN parameters from LLMs, leading to homogenized expert functions and weakening the intended diversification of the MoE architecture. Meanwhile, router rigidity stems from the prevalent use of static linear routers for expert selection, which fail to distinguish between visual and textual tokens, resulting in similar expert distributions for image and text. To address these limitations, we propose EvoMoE, an innovative MoE tuning framework. EvoMoE introduces a meticulously designed expert initialization strategy that progressively evolves multiple robust experts from a single trainable expert, a process termed expert evolution that specifically targets severe expert homogenization. Furthermore, we introduce the Dynamic Token-aware Router (DTR), a novel routing mechanism that allocates input tokens to appropriate experts based on their modality and intrinsic token values. This dynamic routing is facilitated by hypernetworks, which dynamically generate routing weights tailored for each individual token. Extensive experiments demonstrate that EvoMoE significantly outperforms other sparse MLLMs across a variety of multi-modal benchmarks, including MME, MMBench, TextVQA, and POPE. Our results highlight the effectiveness of EvoMoE in enhancing the performance of MLLMs by addressing the critical issues of expert uniformity and router rigidity.

Weakly supervised semantic segmentation (WSSS) aims to provide accurate pixel-level annotation based on only weak guidance, primarily derived from image-level labels. Recent WSSS methods exploit pseudo-labels generated from improved class activation maps (CAMs) to train a fine-grained classification model for semantic segmentation. However, these pseudo-labels are unreliable because they tend to either miss parts of the objects or include irrelevant regions due to weak guidance from individual images. In this paper, we propose a simple yet effective iterative mix-up strategy, Pseudo-Label-based Mix (PL-Mix), that refines pseudo-labels iteratively, thereby further enhancing WSSS performance. During each iteration, we migrate object regions from pseudo-labels produced in previous steps and render them with new contexts in a mix-up fashion. Due to model consistency enforcement across varied backgrounds and new combinations of multiple objects from enriched image samples, these pseudo-labels progressively become more accurate and reliable. Further enhanced by a masking strategy and a CAM-based earth mover’s distance loss, we achieve state-of-the-art performance on the PASCAL VOC2012 and MS COCO2014 benchmark datasets.

Accurate identification of protein nucleic acid binding residues poses a significant challenge with important implications for various biological processes and drug design. Many typical computational methods for protein analysis rely on a single model that could ignore either the semantic context of the protein or the global 3D geometric information. Consequently, these approaches may result in incomplete or inaccurate protein analysis. To address the above issue, in this paper, we present CrossBind, a novel collaborative cross modal approach for identifying binding residues by exploiting both protein geometric structure and its sequence prior knowledge extracted from a large scale protein language model. Specifically, our multi modal approach leverages a contrastive learning technique and atom wise attention to capture the positional relationships between atoms and residues, thereby incorporating fine grained local geometric knowledge, for better binding residue prediction. Extensive experimental results demonstrate that our approach outperforms the next best state of the art methods, GraphSite and GraphBind, on DNA and RNA datasets by 10.8/17.3% in terms of the harmonic mean of precision and recall (F1 Score) and 11.9/24.8% in Matthews correlation coefficient (MCC), respectively. We release the code at https://github.com/BEAM-Labs/CrossBind.

The field of 4D point cloud understanding is rapidly developing with the goal of analyzing dynamic 3D point cloud sequences. However, it remains a challenging task due to the sparsity and lack of texture in point clouds. Moreover, the irregularity of point cloud poses a difficulty in aligning temporal information within video sequences. To address these issues, we propose a novel cross-modal knowledge transfer framework, called X4D-SceneFormer. This framework enhances 4D-Scene understanding by transferring texture priors from RGB sequences using a Transformer architecture with temporal relationship mining. Specifically, the framework is designed with a dual-branch architecture, consisting of an 4D point cloud transformer and a Gradient-aware Image Transformer (GIT). The GIT combines visual texture and temporal correlation features to offer rich semantics and dynamics for better point cloud representation. During training, we employ multiple knowledge transfer techniques, including temporal consistency losses and masked self-attention, to strengthen the knowledge transfer between modalities. This leads to enhanced performance during inference using single-modal 4D point cloud inputs. Extensive experiments demonstrate the superior performance of our framework on various 4D point cloud video understanding tasks, including action recognition, action segmentation and semantic segmentation. The results achieve 1st places, i.e., 85.3% (+7.9%) accuracy and 47.3% (+5.0%) mIoU for 4D action segmentation and semantic segmentation, on the HOI4D challenge, outperforming previous state-of-the-art by a large margin. We release the code at https://github.com/jinglinglingling/X4D.

Event-based semantic segmentation has gained popularity due to its capability to deal with scenarios under high-speed motion and extreme lighting conditions, which cannot be addressed by conventional RGB cameras. Since it is hard to annotate event data, previous approaches rely on event-to-image reconstruction to obtain pseudo labels for training. However, this will inevitably introduce noise, and learning from noisy pseudo labels, especially when generated from a single source, may reinforce the errors. This drawback is also called confirmation bias in pseudo-labeling. In this paper, we propose a novel hybrid pseudo-labeling framework for unsupervised event-based semantic segmentation, HPL-ESS, to alleviate the influence of noisy pseudo labels. Specifically, we first employ a plain unsupervised domain adaptation framework as our baseline, which can generate a set of pseudo labels through self-training. Then, we incorporate offline event-to-image re-construction into the framework, and obtain another set of pseudo labels by predicting segmentation maps on the re-constructed images. A noisy label learning strategy is designed to mix the two sets of pseudo labels and enhance the quality. Moreover, we propose a soft prototypical alignment (SPA) module to further improve the consistency of target domain features. Extensive experiments show that the proposed method outperforms existing state-of-the-art methods by a large margin on benchmarks (e.g., +5.88% accuracy, +10.32% mIoU on DSEC-Semantic dataset), and even surpasses several supervised methods.