DTVI: Dual-Stage Textual and Visual Intervention for Safe Text-to-Image Generation

Dimensions overall score 7.0

• A dual-stage defense framework for safe text-to-image generation that intervenes at both textual and visual stages to capture and attenuate unsafe content. Existing inference-time defense methods typically perform category-agnostic token-level intervention in the text embedding space, which fails to capture malicious semantics distributed across the full token sequence and remains vulnerable to adversarial prompts.problem 35%
  Evidencepartial

  A dual-stage defense framework for safe text-to-image generation that intervenes at both textual and visual stages to capture and attenuate unsafe content. Existing inference-time defense methods typically perform category-agnostic token-level intervention in the text embedding space, which fails to capture malicious semantics distributed across the full token sequence and remains vulnerable to adversarial prompts.

  Implicationpartial

  Abstract-backed fallback claim; anchored extraction has not materialized a public claim row yet.

  Verificationpartial

  partial

• Text-to-Image (T2I) diffusion models have demonstrated strong generation ability, but their potential to generate unsafe content raises significant safety concerns. Existing inference-time defense methods typically perform category-agnostic token-level intervention in the text embedding space, which fails to capture malicious semantics distributed across the full token sequence and remains vulnerable to adversarial prompts.method 35%
  Evidencepartial

  Text-to-Image (T2I) diffusion models have demonstrated strong generation ability, but their potential to generate unsafe content raises significant safety concerns. Existing inference-time defense methods typically perform category-agnostic token-level intervention in the text embedding space, which fails to capture malicious semantics distributed across the full token sequence and remains vulnerable to adversarial prompts.

  Implicationpartial

  Abstract-backed fallback claim; anchored extraction has not materialized a public claim row yet.

  Verificationpartial

  partial

• ScienceToStartup currently rates this 7.0/10 on the public viability pass. Experimental results on real-world unsafe prompts, adversarial prompts, and multiple harmful categories show that our method achieves effective and robust defense while preserving reasonable generation quality on benign prompts, obtaining an average Defense Success Rate (DSR) of 94.43% across sexual-category benchmarks and 88.56 across seven unsafe categories, while maintaining generation quality on benign prompts. Code availability is flagged in the production record; the public repository link still needs proof alignment.result 35%
  Evidencepartial

  ScienceToStartup currently rates this 7.0/10 on the public viability pass. Experimental results on real-world unsafe prompts, adversarial prompts, and multiple harmful categories show that our method achieves effective and robust defense while preserving reasonable generation quality on benign prompts, obtaining an average Defense Success Rate (DSR) of 94.43% across sexual-category benchmarks and 88.56 across seven unsafe categories, while maintaining generation quality on benign prompts. Code availability is flagged in the production record; the public repository link still needs proof alignment.

  Implicationpartial

  Abstract-backed fallback claim; anchored extraction has not materialized a public claim row yet.

  Verificationpartial

  partial

• Generative AI Safety moved forward this cycle; last verified April 2026. Public score 7.0/10. Production flags indicate code availability.why now 35%
  Evidencepartial

  Generative AI Safety moved forward this cycle; last verified April 2026. Public score 7.0/10. Production flags indicate code availability.

  Implicationpartial

  Abstract-backed fallback claim; anchored extraction has not materialized a public claim row yet.

  Verificationpartial

  partial

Competitive landscape

A dual-stage defense framework for safe text-to-image generation that intervenes at both textual and visual stages to capture and attenuate unsafe content.