arXiv:2512. 06276v3 Announce Type: replace-cross Abstract: Referring Expression Comprehension (REC) is a vision-language task that localizes a specific image region based on a textual description.
ContourVLA presents a closed‑loop vision‑language‑action framework for generalized referring expression segmentation. It uses editable contours as policy states, coupled with Evolution‑Aware Semantic Scheduling to adapt perception, and a Dustbin‑Augmented Entropic Credit Transport GRPO to jointly optimize grounding and contour actions. The method achieves significant gains in gIoU and mIoU over existing baselines on multiple RefCOCO datasets.
The paper introduces OpenRef, a benchmark for Referring Expression Comprehension (REC) designed for open‑world scenarios. OpenRef expands beyond simple settings by including diverse visual domains, variable target counts (multi‑target and none‑target), and a rich vocabulary with proper nouns, polysemous words, and ordinal terms. It also proposes new evaluation metrics—F1 for grounding accuracy and N3R for negative expression rejection—and presents a training‑free Multi‑task Consistency Checker (MCC) that improves model performance with a single click.
arXiv:2609.24510v2 Announce Type: replace Abstract: Recent advances in vision foundation models (VFMs) have shown remarkable capabilities across diverse unimodal visual tasks. However, adapting VFMs...
The paper introduces Bidirectional Reciprocal Learning (BRL), a parameter‑efficient fine‑tuning framework for referring image segmentation that operates on frozen vision foundation models. BRL employs two lightweight adapters—Reciprocal Attention Adapter (RAA) for token‑level cross‑modal attention and Reciprocal Gate Adapter (RGA) for channel‑level gating—to enable hierarchical, bidirectional information flow between vision and language. Experiments on RefCOCO, RefCOCO+, and RefCOCOg show that BRL outperforms existing methods while updating fewer than 0.5% of backbone parameters.
arXiv:2607.10744v5 Announce Type: replace Abstract: Benefiting from the powerful priors embedded in large-scale pre-training data and the emerging commonsense reasoning ability, large language models...
arXiv:2610.07720v1 Announce Type: cross Abstract: Multi-reference image generation requires preserving the appearance of multiple subjects while composing them into a coherent scene. However, existin...
DRAgent introduces a discriminative reasoning framework for Referring Expression Segmentation that replaces direct coordinate prediction with a two‑stage selection process. First, a detector generates candidate boxes; then a multimodal large language model (MLLM) discriminates the correct target through high‑recall screening and instance‑wise verification. The chosen box serves as a spatial prompt for a foundation segmentation model, and a self‑consistency‑filtered data pipeline is used for LoRA‑based fine‑tuning, yielding competitive results on RefCOCO, RefCOCO+, and RefCOCOg.
arXiv:2606. 17924v1 Announce Type: cross Abstract: Current Vision-Language-Action (VLA) models face a trade-off between efficient action generation and explicit deliberation.
Referring Expression Segmentation (RES) aims to generate a pixel-level mask for the object specified by a language expression. Recent methods based on multimodal large language models (MLLMs) often re...
arXiv:2608.29126v1 Announce Type: new Abstract: Referring single-object tracking enables language-grounded target initialization and subsequent tracking by jointly leveraging semantic cues and visual...
The paper introduces DRMOT, a new dataset and framework for Referring Multi-Object Tracking (RMOT) that incorporates RGB-D data. DRSet contains 2,076 referring expressions across 186 RGB-D sequences, pairing synchronized RGB images, depth maps, and language descriptions—including depth-dependent cues—to evaluate spatial‑semantic detection and tracking. The proposed DRTrack framework uses a Geometric‑Aware GRPO fine‑tuned Multimodal Large Language Model to perform precise detection and robust trajectory association from joint language, RGB, and depth inputs, with experiments demonstrating its effectiveness.