REVIEW 7 cited by
Compile Scene Graphs with Reinforcement Learning
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
read the original abstract
Next-token prediction is the fundamental principle for training large language models (LLMs), and reinforcement learning (RL) further enhances their reasoning performance. As an effective way to model language, image, video, and other modalities, the use of LLMs for end-to-end extraction of structured visual representations, such as scene graphs, remains underexplored. It requires the model to accurately produce a set of objects and relationship triplets, rather than generating text token by token. To achieve this, we introduce R1-SGG, a multimodal LLM (M-LLM) initially trained via supervised fine-tuning (SFT) on the scene graph dataset and subsequently refined using reinforcement learning to enhance its ability to generate scene graphs in an end-to-end manner. The SFT follows a conventional prompt-response paradigm, while RL requires the design of effective reward signals. We design a set of graph-centric rewards, including three recall-based variants -- Hard Recall, Hard Recall+Relax, and Soft Recall -- which evaluate semantic and spatial alignment between predictions and ground truth at the object and relation levels. A format consistency reward further ensures that outputs follow the expected structural schema. Extensive experiments on the VG150 and PSG benchmarks show that R1-SGG substantially reduces failure rates and achieves strong performance in Recall and mean Recall, surpassing traditional SGG models and existing multimodal language models. Our code is available at https://github.com/gpt4vision/R1-SGG
Forward citations
Cited by 7 Pith papers
-
SenBen: Sensitive Scene Graphs for Explainable Content Moderation
SenBen is the first large-scale scene graph benchmark for sensitive content, paired with a 241M distilled model that outperforms most VLMs and safety APIs on grounded detection while running much faster.
-
SceneGraphVLM: Dynamic Scene Graph Generation from Video with Vision-Language Models
SceneGraphVLM generates dynamic scene graphs from video using compact VLMs, TOON serialization, and hallucination-aware RL to improve precision and achieve one-second latency.
-
SenBen: Sensitive Scene Graphs for Explainable Content Moderation
A 241M multi-task student trained with suffix identity, VAR loss, and a decoupled Q2L head matches or beats most VLMs and safety APIs on grounded sensitive scene graphs at 7.6× lower latency.
-
SpatialThinker: Reinforcing Scene Graph-Grounded Spatial Reasoning via Dense Rewards
Dense scene-graph-grounded rewards let a 7B multimodal LLM trained on 7K synthetic questions beat SFT and sparse-RL baselines and outscore GPT-4o on average across 12 spatial/real-world benchmarks.
-
3D-R1: Enhancing Reasoning in 3D VLMs for Unified Scene Understanding
3D-R1 uses a synthetic chain-of-thought cold start plus GRPO reinforcement learning with perception, semantic, and format rewards, and reports best published results across 3D dense captioning, QA, grounding, dialogue...
-
OracleAnalyser: Analysing Implicit Semantics of Oracle Bone Scripts through MLLMs with Post-training
OracleAnalyser applies post-training and a new Stable Focal Preference Optimization algorithm to a 3B MLLM for oracle bone script analysis, releasing datasets and a benchmark where the small model outperforms larger ones.
-
Reinforcement Learning Meets Large Language Models: A Survey of Advancements and Applications Across the LLM Lifecycle
A survey that maps reinforcement learning methods, datasets, benchmarks, and open-source tools across the full training lifecycle of large language models, focusing on verifiable-reward reasoning.
Discussion (0). Sign in to comment.