(CVPR 2026)Spatial-SSRL: Enhancing Spatial Understanding via Self-Supervised Reinforcement Learning

Yuhong Liu · Beichen Zhang · Yuhang Zang^† · Yuhang Cao · Long Xing
Xiaoyi Dong · Haodong Duan · Dahua Lin · Jiaqi Wang^†

^†Corresponding authors.

📖Paper | 🏠Homepage | 🤗Spatial-SSRL-7B Model | 🤗Spatial-SSRL-3B Model | 🤗Spatial-SSRL-Qwen3VL-4B Model | 🤗Spatial-SSRL-81k Dataset | 📰Daily Paper

📢 News

• 🚀 [2026/04/05] We have released the training code of Spatial-SSRL.
• 🚀 [2026/02/25] We have released the 🤗Spatial-SSRL-3B Model, initialized from Qwen2.5-VL-3B-Instruct.
• 🚀 [2026/02/21] Our work has been accepted by CVPR 2026.
• 🚀 [2025/11/24] We have released the 🤗Spatial-SSRL-Qwen3VL-4B Model, initialized from Qwen3-VL-4B-Instruct.
• 🚀 [2025/11/03] Now you can try out Spatial-SSRL-7B on 🤗Spatial-SSRL Space.
• 🚀 [2025/11/03] We have released the Spatial-SSRL-7B Model, and Spatial-SSRL-81k Dataset.
• 🚀 [2025/11/02] We have released the Spatial-SSRL repository.

🌈 Overview

We are thrilled to introduce Spatial-SSRL, a novel self-supervised RL paradigm aimed at enhancing LVLM spatial understanding. By optimizing Qwen2.5-VL-7B with Spatial-SSRL, the model exhibits stronger spatial intelligence across seven spatial understanding benchmarks in both image and video settings.

Spatial-SSRL is a lightweight tool-free framework that is natually compatible with the RLVR training paradigm and easy to extend to a multitude of pretext tasks. Five tasks are currently formulated in the framework, requiring only ordinary RGB and RGB-D images. And we welcome you to join Spatial-SSRL with effective pretext tasks to further strengthen the capabilities of LVLMs!

💡 Highlights

• 🔥 Highly Scalable: Spatial-SSRL uses ordinary raw RGB and RGB-D images instead of richly-annotated public datasets or manual labels for data curation, making it highly scalable.
• 🔥 Cost-effective: Avoiding the need for human labels or API calls for general LVLMs throughout the entire pipeline endows Spatial-SSRL with cost-effectiveness.
• 🔥 Lightweight: Prior approaches for spatial understanding heavily rely on annotation of external tools, incurring inherent errors in training data and additional cost. In constrast, Spatial-SSRL is completely tool-free and can easily be extended to more self-supervised tasks.
• 🔥 Naturally Verifiable: Intrinsic supervisory signals determined by pretext objectives are naturally verifiable, aligning Spatial-SSRL well with the RLVR paradigm.

📊 Results

We train Qwen2.5-VL-3B and Qwen2.5-VL-7B with our Spatial-SSRL paradigm and the experimental results across seven spatial understanding benchmarks are shown below.

⭐️ Quick Start

To directly experience Spatial-SSRL-7B, you can try it out on 🤗Spatial-SSRL Space!

Here we provide a code snippet for you to start a simple trial of Spatial-SSRL-7B on your own device. You can download the model from 🤗Spatial-SSRL-7B Model before your trial!

fromtransformersimportQwen2_5_VLForConditionalGeneration, AutoTokenizer, AutoProcessorfromqwen_vl_utilsimportprocess_vision_infomodel_path="internlm/Spatial-SSRL-7B"#You can change it to your own local path if deployed alreadyimg_path="examples/eg1.jpg"question="Consider the real-world 3D locations of the objects. Which object has a higher location? A. yellow bear kite B. building"#We recommend using the format prompt to make the inference consistent with trainingformat_prompt="\n You FIRST think about the reasoning process as an internal monologue and then provide the final answer. The reasoning process MUST BE enclosed within <think> </think> tags. The final answer MUST BE put in \\boxed{}."model=Qwen2_5_VLForConditionalGeneration.from_pretrained(
model_path, torch_dtype="auto", device_map="auto"
)
processor=AutoProcessor.from_pretrained(model_path)
messages= [
{
"role": "user",
"content": [
{
"type": "image",
"image": img_path,
},
{"type": "text", "text": question+format_prompt},
],
}
]
text=processor.apply_chat_template(
messages, tokenize=False, add_generation_prompt=True
)
image_inputs, video_inputs=process_vision_info(messages)
inputs=processor(
text=[text],
images=image_inputs,
videos=video_inputs,
padding=True,
return_tensors="pt",
)
inputs=inputs.to("cuda")
generated_ids=model.generate(**inputs, max_new_tokens=4096, do_sample=False)
generated_ids_trimmed= [
out_ids[len(in_ids) :] forin_ids, out_idsinzip(inputs.input_ids, generated_ids)
]
output_text=processor.batch_decode(
generated_ids_trimmed, skip_special_tokens=True, clean_up_tokenization_spaces=False
)
print("Model Response:", output_text)

Here we provide a code snippet for you to start a simple trial of Spatial-SSRL-Qwen3VL-4B on your own device. You can download the model from 🤗Spatial-SSRL-Qwen3VL-4B Model before your trial!

fromtransformersimportAutoProcessor, AutoModelForImageTextToText#transformers==4.57.1fromqwen_vl_utilsimportprocess_vision_info#0.0.14importtorchmodel_path="internlm/Spatial-SSRL-Qwen3VL-4B"#You can change it to your own local path if deployed already#Change the path of the input imageimg_path="examples/eg_qwen3vl.jpg"#Change your question herequestion="Question: Consider the real-world 3D locations and orientations of the objects. If I stand at the man's position facing where it is facing, is the menu on the left or right of me?\nOptions:\nA. on the left\nB. on the right\n"question+="Please select the correct answer from the options above. \n"#We recommend using the format prompt to make the inference consistent with trainingformat_prompt="You FIRST think about the reasoning process as an internal monologue and then provide the final answer. The reasoning process MUST BE enclosed within <think> </think> tags. The final answer MUST BE put in \\boxed{}."model=AutoModelForImageTextToText.from_pretrained(
model_path, torch_dtype=torch.float16, device_map='auto', attn_implementation='flash_attention_2'
)
processor=AutoProcessor.from_pretrained(model_path)
messages= [
{
"role": "user",
"content": [
{
"type": "image",
"image": img_path,
},
{"type": "text", "text": question+format_prompt},
],
}
]
text=processor.apply_chat_template(
messages, tokenize=False, add_generation_prompt=True
)
image_inputs, video_inputs=process_vision_info(messages)
inputs=processor(
text=[text],
images=image_inputs,
videos=video_inputs,
padding=True,
return_tensors="pt",
)
inputs=inputs.to("cuda")
generated_ids=model.generate(**inputs, max_new_tokens=4096, do_sample=False)
generated_ids_trimmed= [
out_ids[len(in_ids) :] forin_ids, out_idsinzip(inputs.input_ids, generated_ids)
]
output_text=processor.batch_decode(
generated_ids_trimmed, skip_special_tokens=True, clean_up_tokenization_spaces=False
)
print("Model Response:", output_text[0])

🛠️ Training

See training.md

🛠️ Evaluation

Prepare your environment:

git clone https://github.com/InternLM/Spatial-SSRL.git
conda create -n spatialssrl python==3.10
conda activate spatialssrl
cd Spatial-SSRL/evaluation
pip install -r requirements.txt
# Recommended
pip install flash-attn --no-build-isolation

Start your evaluation by referring to the tutorials in Eval.md

Cases

✒️Citation

If you find this project useful, please kindly cite:

@article{liu2025spatial,
title={Spatial-SSRL: Enhancing Spatial Understanding via Self-Supervised Reinforcement Learning},
author={Liu, Yuhong and Zhang, Beichen and Zang, Yuhang and Cao, Yuhang and Xing, Long and Dong, Xiaoyi and Duan, Haodong and Lin, Dahua and Wang, Jiaqi},
journal={arXiv preprint arXiv:2510.27606},
year={2025}
}

📄 License

Usage and License Notices: The data and code are intended and licensed for research use only.

❤️ Acknowledgement

We extend our sincere gratitude to VLMEvalkit, the powerful toolkit to evaluate a vast range of LMMs!

The work is built upon EasyR1 and LlaMaFactory, two excellent codebase for LLM training!