菜单
Enforce semantic compatibility constraints directly into instance construction for 3D object detection. Uses MLLM-driven semantic grouping to partition object vocabulary into plausible cross-view category equivalence clusters, preventing geometry-driven over-merging.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Convert arXiv papers into ready-to-use agent skills using category-aware extraction. First classifies the paper into one or more of 11 research categories, then applies a specialized extraction pipeline for each category — because different types of papers produce different types of usable knowledge. A single paper can yield multiple skills if it spans categories. Use this skill whenever the user wants to turn a paper into a skill, extract practical techniques from research, build a skill library from papers, convert arXiv papers into reusable agent instructions, or batch-process multiple papers into skills. Also trigger when someone asks about extracting actionable knowledge from papers, making research practical for LLM agents, or systematically converting academic contributions into structured agent capabilities.
Establish regret bounds for behavior cloning with discretized actions combining statistical error and quantization error terms. Prove smoothness requirements for safe quantizer design, show that learning-based quantizers fail these requirements, and propose model-based augmentation to reduce error dependence from H² to H.
Dynamically allocate LoRA ranks per-layer during fine-tuning instead of using fixed uniform ranks. Learn optimal rank for each layer and subject via variational framework with discretized exponential distribution, reducing memory footprint while maintaining fidelity and text-alignment.
Adapt general LLMs to specialized manufacturing domains via domain-adaptive pretraining on open-access journals and visual instruction tuning. Extract 50M tokens and 24K images from peer-reviewed papers, achieve >90% accuracy on domain knowledge tasks, and enable real-time defect identification from manufacturing images.
Future intelligence explosions will be plural, social, and entangled with humanity through distributed collaborative systems rather than singular superintelligence. Intelligence is inherently social, demanding infrastructure matching agent development; integrate governance, institutional frameworks, and constitutional checks across hierarchies of autonomous agents and human-AI centaurs in shifting configurations.
Build taxonomy-aware audio-text pretraining systems for species recognition from animal vocalizations. Train contrastive models that augment text prompts with hierarchical taxonomic structure (scientific/common names, phylogenetic sequences), evaluate on unseen species via rare-species test sets, and predict ecological traits directly from audio.
| name | group3d-semantic-grouping-detection |
| title | Group3D: MLLM-Driven Semantic Grouping for Open-Vocabulary 3D Object Detection |
| version | 0.0.3 |
| engine | skillxiv-v0.0.3-claude-opus-4.6 |
| license | MIT |
| url | https://arxiv.org/abs/2603.21944 |
| keywords | ["3D Detection","Semantic Grouping","MLLM","Multi-view","Open-Vocabulary"] |
| description | Enforce semantic compatibility constraints directly into instance construction for 3D object detection. Uses MLLM-driven semantic grouping to partition object vocabulary into plausible cross-view category equivalence clusters, preventing geometry-driven over-merging. |
Group-Gated Fragment Merging with Semantic Compatibility Constraints
Open-vocabulary 3D detection from multi-view RGB images typically decouples instance construction from semantic labeling. Fragments are merged based on geometric consistency first, then labeled. This causes irreversible errors when geometric evidence is incomplete or view-dependent, and the approach cannot recover from geometry-driven over-merging decisions.
Group3D integrates semantic constraints into the merge decision itself, requiring both semantic compatibility and geometric consistency before combining fragments:
Scene-Adaptive Vocabulary Memory - Query a multimodal large language model (MLLM) across views to identify object categories present in the scene, then aggregate into a scene-specific vocabulary.
Semantic Compatibility Groups - The MLLM partitions the vocabulary into groups capturing "plausible cross-view category equivalence" under taxonomy noise. Groups identify categories that could refer to the same physical object while excluding structurally incompatible associations.
Group-Gated Fragment Merging - Two 3D fragments merge only when they satisfy both conditions:
The algorithm processes fragments sorted by spatial extent, attempting merges with existing clusters while respecting both gates.
The paper demonstrates:
