English static mirror for SEO/GEO · AI-assisted translation · Read Chinese original

LocateAnything: Parallel Box Decoding for Fast, High-Quality VLM Grounding

Forum topic · 小凯 · 2026-06-07

Summary

LocateAnything, a vision-language grounding model from NVIDIA and collaborators (arXiv:2605.27365), replaces sequential token-by-token bounding box generation with Parallel Box Decoding (PBD). Instead of serializing boxes into 1D token streams, PBD treats each box as an atomic 6-token block generated in a single forward pass. The model is trained jointly on next-token-prediction (NTP) and multi-token-prediction (MTP) streams with a carefully designed attention mask combining causal, block-causal, and intra-block bidirectional attention. It offers three inference modes: Fast (16.9 boxes/sec), Slow (3.9 BPS), and a Hybrid mode (12.7 BPS) that falls back to token-level decoding when format anomalies or spatial ambiguity are detected. Training uses LocateAnything-Data, covering 12M images, 138M language queries, and 785M boxes across detection, referring, UI, document, OCR, and point localization tasks, with automatic query synthesis and explicit negative samples. The 3B model achieves ~10× higher throughput than Qwen3-VL-4B, +3.8% LVIS mean F1, +4.1% VisDrone, +14.5% M6Doc, and 60.3 F1 on ScreenSpot-Pro, outperforming 30B+ models.

Overview

Paper: *LocateAnything: Fast and High-Quality Vision-Language Grounding with Parallel Box Decoding* Authors: Shihao Wang, Shilong Liu, Yuanguo Kuang, Xinyu Wei, Yangzhou Liu, Zhiqi Li, Yunze Man, Guo Chen, Andrew Tao, Guilin Liu, Jan Kautz, Lei Zhang, Zhiding Yu (NVIDIA et al.) Link: https://arxiv.org/abs/2605.27365

Headline results: 2.5× decoding speedup, LVIS mean F1 +3.8%, and a unified 138M-sample training set covering detection, grounding, UI, documents, and OCR.

The Problem: Sequential "Toothpaste" Decoding

Mainstream VLMs serialize 2D bounding boxes into 1D token streams—either as digit characters ("1" "0" "2" "4" ...) or quantized coordinates predicted one at a time. This causes:

  • Inference bottleneck: each coordinate waits for the previous one (strictly serial)
  • Structural mismatch: box coordinates are inherently coupled, but tokens are trained largely independently
  • Error propagation: one wrong coordinate skews the whole box
  • Low throughput: 300 boxes = 1200 serial steps in dense scenes
  • Core Idea: Parallel Box Decoding (PBD)

    Since a bounding box is a geometric whole, LocateAnything generates it as an atomic unit. Four fixed-length (6-token) block types are defined:

    | Block | Content | Length | |---|---|---| | Semantic | language description (long text split into blocks) | 6 | | Box | <box> x1 y1 x2 y2 </box> | 6 | | Negative | target absent | 6 | | End | termination | 6 |

    All 6 tokens of a box block are generated in parallel in one forward pass.

    Three Inference Modes

    | Mode | Mechanism | Throughput | Use case | |---|---|---|---| | Fast | pure MTP, boxes in parallel | 16.9 BPS | latency-sensitive, robotics, embedded | | Slow | pure NTP, token-by-token | 3.9 BPS | high-precision annotation | | Hybrid | Fast by default, fallback to Slow | 12.7 BPS | production, balanced |

    Hybrid mode falls back on: (1) format anomalies (token order deviates from the box structure), or (2) spatial ambiguity (top-1 coordinate probability < 0.7 and top-5 max-min spread > 80 in the 0–1000 normalized space).

    Training: Dual-Sequence Joint Supervision

    Input: x_all = [visual_tokens] + [query_tokens] + [NTP_sequence] + [MTP_sequence]

    The attention mask has three regimes: 1. Causal attention over the NTP stream + shared context 2. Block-causal between MTP blocks 3. Bidirectional intra-block within each 6-token MTP block

    NTP and MTP streams cannot see each other (no data leakage), but both see the shared visual/text context. Loss: L = L_NTP + L_MTP. Ablations show joint training is essential—pure MTP training collapses to 47.2 F1 vs 52.1 for joint (Slow).

    Data Engine: LocateAnything-Data

  • 12M unique images, 138M natural-language queries, 785M boxes
  • Task mix: general detection 66.9%, UI grounding 16.5%, referring 7.3%, text localization 3.6%, document/layout 3.5%, point localization 2.2%
  • Synthesis pipeline: detection datasets (OpenImages, Objects365) are augmented by prompting Qwen3-VL with class labels to generate rich queries; Molmo predicts points, and points inside ground-truth boxes are kept as supervision. For unlabeled images, Qwen3-VL generates queries, Molmo points are converted to boxes with SAM 3, then Qwen3-VL verifies.
  • Explicit negative samples teach the model to say "not found" instead of hallucinating.
  • Results

    Throughput (H100, batch size 1)

    | Model | BPS | Relative | |---|---|---| | Qwen3-VL-4B | 1.1 | 1× | | Rex-Omni-3B | 5.0 | 4.5× | | LocateAnything-3B | 12.7 | 11.5× |

    Accuracy

    | Benchmark | Rex-Omni-3B | LocateAnything-3B | |---|---|---| | LVIS mean F1 | 46.9 | 50.7 (+3.8) | | COCO mean F1 | 52.9 | 54.7 (+1.8) | | Dense200 mean F1 | 58.3 | 58.7 | | VisDrone mean F1 | 35.8 | 39.9 (+4.1) | | ScreenSpot-Pro F1 | (GUI-Owl-32B: 58.0) | 60.3 | | DocLayNet mean F1 | 70.7 | 76.8 (+6.1) | | M6Doc mean F1 | 55.6 | 70.1 (+14.5) | | TotalText mean F1 | 40.6 | 43.3 (+2.7) | | RefCOCOg val mean F1 | 73.6 | 76.7 (+3.1) |

    Ablations

  • Coordinate representation: box-aligned PBD beats textual (1.3 BPS, 49.1 F1) and quantized (3.9 BPS, 50.1 F1) representations; PBD (Slow) reaches 52.1 F1 at the same throughput.
  • MTP formulation: structure-agnostic multi-token prediction (SDLM, Block Diff) is harmful (44.8–46.5 F1); box-aligned PBD (Fast) reaches 49.6 F1 at 16.9 BPS.
  • Loss: joint NTP+MTP is required; MTP-only training underperforms.
  • Why It Matters

  • Breaks the "generation = serial" assumption: structured outputs (3D boxes, keypoints, masks, trajectories) could likewise be generated as atomic parallel units.
  • Data-scale effect: 3B parameters beat 30B+ models on several benchmarks.
  • Embodied AI: Fast Mode (16.9 BPS) enables real-time perception; Hybrid Mode preserves reliability.
  • Engineering highlights: Stream Packing (>95% batch utilization for variable-length sequences) and MagiAttention (distributed training with heterogeneous attention masks).
  • Limitations

  • SFT only; RL-based optimization of block-level policies is future work.
  • Validated on boxes and points only; polygons, 3D cuboids, and poses remain untested.
  • Training cost: 4 stages, 256×H100, tens of thousands of steps.
  • Conclusion

    LocateAnything elevates the bounding box from a 1D token sequence to an atomic parallel unit, addressing speed (10× throughput), precision (box-aligned supervision), and robustness (Hybrid fallback). Combined with the 138M-sample data engine, it offers a complete unified visual grounding solution—a statement that multimodal generation's future lies in smarter structures, not longer sequences.

    References

  • Wang S, Liu S, Kuang Y, et al. LocateAnything: Fast and High-Quality Vision-Language Grounding with Parallel Box Decoding. arXiv:2605.27365, 2026.
  • Bai K, et al. Qwen3-VL Technical Report. 2025.
  • Jiang Y, et al. Rex-Omni: Unified Detection and Grounding in a VLM Framework. 2025.
  • Chen J, et al. Pix2Seq: A Language Modeling Framework for Object Detection. ICLR, 2022.

Tags

#locatanything#parallel-box-decoding#vlm#visual-grounding#object-detection#nvidia#multimodal#ai-efficiency

This page is an English static mirror generated for search and AI citation. It may be a full translation or structured summary of the Chinese original. Canonical interactive discussion lives on the Chinese page: https://zhichai.net/topic/177980924