JanusCoder: Towards a Foundational Visual-Programmatic Interface for Code Intelligence
By: Qiushi Sun , Jingyang Gong , Yang Liu and more
Potential Business Impact:
Makes computers create pictures from code.
The scope of neural code intelligence is rapidly expanding beyond text-based source code to encompass the rich visual outputs that programs generate. This visual dimension is critical for advanced applications like flexible content generation and precise, program-driven editing of visualizations. However, progress has been impeded by the scarcity of high-quality multimodal code data, a bottleneck stemming from challenges in synthesis and quality assessment. To address these challenges, we make contributions from both a data and modeling perspective. We first introduce a complete synthesis toolkit that leverages reciprocal synergies between data modalities to efficiently produce a large-scale, high-quality corpus spanning from standard charts to complex interactive web UIs and code-driven animations. Leveraging this toolkit, we construct JanusCode-800K, the largest multimodal code corpus to date. This powers the training of our models, JanusCoder and JanusCoderV, which establish a visual-programmatic interface for generating code from textual instructions, visual inputs, or a combination of both. Our unified model is a departure from existing approaches that build specialized models for isolated tasks. Extensive experiments on both text-centric and vision-centric coding tasks demonstrate the superior performance of the JanusCoder series, with our 7B to 14B scale models approaching or even exceeding the performance of commercial models. Furthermore, extensive analysis provides key insights into harmonizing programmatic logic with its visual expression. Our code and checkpoints will are available at https://github.com/InternLM/JanusCoder.
Similar Papers
Janus-Pro: Unified Multimodal Understanding and Generation with Data and Model Scaling
Artificial Intelligence
Makes computers create better pictures from words.
ScreenCoder: Advancing Visual-to-Code Generation for Front-End Automation via Modular Multimodal Agents
CV and Pattern Recognition
Turns screen designs into working computer code.
VisCodex: Unified Multimodal Code Generation via Merging Vision and Coding Models
Computation and Language
Helps computers write code from pictures.