Large vision-language models (LVLMs) have significantly advanced numerous fields. In this work, we explore how to harness their potential to address 3D scene understanding tasks, using 3D question answering (3D-QA) as a representative example. Due to the limited training data in 3D, we do not train LVLMs but infer in a zero-shot manner. Specifically, we sample 2D views from a 3D point cloud and feed them into 2D models to answer a given question. When the 2D model is chosen, e.g., LLAVA-OV, the quality of sampled views matters the most. We propose cdViews, a novel approach to automatically selecting critical and diverse Views for 3D-QA. cdViews consists of two key components: viewSelector prioritizing critical views based on their potential to provide answer-specific information, and viewNMS enhancing diversity by removing redundant views based on spatial overlap. We evaluate cdViews on the widely-used ScanQA and SQA benchmarks, demonstrating that it achieves state-of-the-art performance in 3D-QA while relying solely on 2D models without fine-tuning. These findings support our belief that 2D LVLMs are currently the most effective alternative (of the resource-intensive 3D LVLMs) for addressing 3D tasks.

The detection of shiitake mushroom fruiting bodies under the cultivation mode of mushroom logs is a critical step in factory-based production. However, existing methods primarily rely on RGB images for feature extraction, which limits the perception of spatial variations such as growth positions and height differences on the logs, leading to insufficient location information and reduced detection accuracy. To address these limitations, this paper proposes a lightweight multi-modal feature fusion detection network (LMFD) based on YOLOv11. First, a dual-backbone architecture is constructed to independently extract color and spatial features, effectively mitigating feature interference caused by discrepancies between modalities. To reduce the computational overhead introduced by the dual backbone, a lightweight feature extraction module named C3k2 Partial Convolution (C3k2P) is designed by replacing standard convolutions with partial convolutions, significantly decreasing the number of parameters and computational complexity. Furthermore, a Cross-domain Dual Attention Fusion (CDAF) module is introduced, which combines spatial and channel attention mechanisms to adaptively enhance multi-modal feature interaction, thereby mitigating feature imbalance and reducing interference caused by depth image noise and voids.Finally, two shiitake mushroom image datasets were constructed for experimentation. Results on the self-built multi-modal dataset (MD) demonstrate that LMFD achieves notable performance improvements, with the mean Average Precision (mAP50) increasing from 87.8% to 91.1%, while maintaining only 12.70M parameters and 27.0 GFLOPs. These results confirm the effectiveness of the proposed model for shiitake mushroom fruiting body detection.

Semantic scene completion (SSC) aims to predict complete 3D voxel occupancy and semantics from a single-view RGB-D image, and recent SSC methods commonly adopt multi-modal inputs. However, our investigation reveals two limitations: ineffective feature learning from single modalities and overfitting to limited datasets. To address these issues, this paper proposes a novel SSC framework - Adversarial Modality Modulation Network (AMMNet) - with a fresh perspective of optimizing gradient updates. The proposed AMMNet introduces two core modules: a cross-modal modulation enabling the interdependence of gradient flows between modalities, and a customized adversarial training scheme leveraging dynamic gradient competition. Specifically, the cross-modal modulation adaptively re-calibrates the features to better excite representation potentials from each single modality. The adversarial training employs a minimax game of evolving gradients, with customized guidance to strengthen the generator's perception of visual fidelity from both geometric completeness and semantic correctness. Extensive experimental results demonstrate that AMMNet outperforms state-of-the-art SSC methods by a large margin, providing a promising direction for improving the effectiveness and generalization of SSC methods. Our code is available at this link.

Semantic Scene Completion (SSC) transforms an image of single-view depth and/or RGB 2D pixels into 3D voxels, each of whose semantic labels are predicted. SSC is a well-known ill-posed problem as the prediction model has to “imagine” what is behind the visible surface, which is usually represented by Truncated Signed Distance Function (TSDF). Due to the sensory imperfection of the depth camera, most existing methods based on the noisy TSDF estimated from depth values suffer from 1) incomplete volumetric predictions and 2) confused semantic labels. To this end, we use the ground-truth 3D voxels to generate a perfect visible surface, called TSDF-CAD, and then train a “cleaner” SSC model. As the model is noise-free, it is expected to focus more on the “imagination” of unseen voxels. Then, we propose to distill the intermediate “cleaner” knowledge into another model with noisy TSDF input. In particular, we use the 3D occupancy feature and the semantic relations of the “cleaner self” to supervise the counterparts of the “noisy self” to respectively address the above two incorrect predictions. Experimental results validate that our method improves the noisy counterparts with 3.1% IoU and 2.2% mIoU for measuring scene completion and SSC, and also achieves new state-of-the-art accuracy on the popular NYU dataset. The code is available at https://github.com/fereenwong/CleanerS.

How to explore useful information from depth is the key success of the RGB-D saliency detection methods. While the RGB and depth images are from different domains, a modality gap will lead to unsatisfactory results for simple feature concatenation. Towards better performance, most methods focus on bridging this gap and designing different cross-modal fusion modules for features, while ignoring explicitly extracting some useful consistent information from them. To overcome this problem, we develop a simple yet effective RGB-D saliency detection method by learning discriminative cross-modality features based on the deep neural network. The proposed method first learns modality-specific features for RGB and depth inputs. And then we separately calculate the correlations of every pixel-pair in a cross-modality consistent way, i.e., the distribution ranges are consistent for the correlations calculated based on features extracted from RGB (RGB correlation) or depth inputs (depth correlation). From different perspectives, color or spatial, the RGB and depth correlations end up at the same point to depict how tightly each pixel-pair is related. Secondly, to complemently gather RGB and depth information, we propose a novel correlation-fusion to fuse RGB and depth correlations, resulting in a cross-modality correlation. Finally, the features are refined with both long-range cross-modality correlations and local depth correlations to predict salient maps. In which, the long-range cross-modality correlation provides context information for accurate localization, and the local depth correlation keeps good subtle structures for fine segmentation. In addition, a lightweight DepthNet is designed for efficient depth feature extraction. We solve the proposed network in an end-to-end manner. Both quantitative and qualitative experimental results demonstrate the proposed algorithm achieves favorable performance against state-of-the-art methods.

Confusing visual appearance and scattered small-scale patterns commonly exist in natural images, which forms a challenge for prior saliency detection methods. Inspired by the sensitivity to edge information of Human Visual Systems, we propose a universal edge-oriented framework to improve the performance of existing salient detection methods. Firstly, edge probability map is extracted from images and utilized to get edge-based over segmentation. Secondly, merging segments by a hierarchical model to generate edge regions. Finally, the proposed framework turns saliency detection to assign a saliency value to each edge region. Experimental results demonstrate the effectiveness of our framework.