- π Model Overview (λͺ¨λΈ κ°μ)
- π‘οΈ Defense & Performance Benchmarks (μ€μ λ°©μ΄ μ§ν)
- β‘ Inference Speed & Runtime Benchmark (μΆλ‘ λ²€μΉλ§ν¬)
- π¦ Export Formats & Deployment Artifacts (λ°°ν¬ νμΌ κ΅¬μ±)
- π License & Commercial Terms (λΌμ΄μ μ€ λ° μμ
μ λμ
μλ΄)
- β οΈ Notes & Disclaimer (κΈ°μ κ³ μ§ λ° μλ΄)
- πΌ Opportunities & Contact (μ±μ© μ μ λ° ννΈλμ)
BAAR3-Captcha-Guard (AI-Resistant Multimodal CAPTCHA Engine)
π Official Website: https://baar.uk/model/
π Model Hub: aixk/BAAR3-Captcha-Guard
π License: Non-Commercial Only (Contact for Commercial Pricing)
π Model Overview (λͺ¨λΈ κ°μ)
BAAR3-Captcha-Guard is an ultra-fast, edge-native multimodal CAPTCHA defense engine designed to counter Vision LLMs (GPT-4o, Claude Computer-Use), automated headless scripts (Puppeteer, Playwright), and replay injection attacks. By fusing visual challenge canvases with user interaction dynamics, the model delivers zero-trust bot discrimination with sub-millisecond edge and WASM latency.
BAAR3-Captcha-Guardλ μ΅μ Vision AI μμ΄μ νΈ(GPT-4o, Claude Computer-Use λ±), λΈλΌμ°μ μλν μ€ν¬λ¦½νΈ(Puppeteer/Selenium), 리νλ μ΄ λ³μ‘° 곡격μ μ°¨λ¨νκΈ° μν μ΄κ²½λ μ£μ§ μΈλΌμΈ 보μ μμ§μ λλ€. μΊλ²μ€ μ΄λ―Έμ§ ν¨ν΄κ³Ό μ¬μ©μ μ‘°μ λμνμ κ²°ν© νλ³νμ¬ μ μ μ μ κ²½νμ ν΄μΉμ§ μμΌλ©΄μ 0.4ms λ΄μΈμ κ·Ήμ μ§μ° νκ²½μμ μ§λ₯ν λ΄μ μ€μκ° μλ³Β·μ°¨λ¨ν©λλ€.
π‘οΈ Defense & Performance Benchmarks (μ€μ λ°©μ΄ μ§ν)
Validation Setup: 10,000 Verified Test Vectors | Runtime: Edge INT8 Quantized Engine
| Class / Category | Classification Type | Detection Rate (μ°¨λ¨μ¨ / ν΅κ³Όμ¨) | Error Metric (μ€λ₯μ¨) | Primary Threat Scenarios |
|---|---|---|---|---|
| Class 0: μ μ μΈκ° (Human) | Benign User | 99.82% ν΅κ³Όμ¨ | FPR: 0.18% |
μμ°μ€λ¬μ΄ μΈμ²΄ μ‘°μ κΆ€μ , μ μ ν°μΉ/λ§μ°μ€ μΈν°λμ |
| Class 1: λ§€ν¬λ‘/μ€ν¬λ¦½νΈ (Macro) | Threat Flow | 100.00% μ°¨λ¨μ¨ | FNR: 0.00% |
Puppeteer/Playwright μλν, λ±μλ λ° λ² μ§μ΄ μ€ν¬λ¦½νΈ |
| Class 2: Vision AI (LLM Agent) | Threat Flow | 99.88% μ°¨λ¨μ¨ | FNR: 0.12% |
λ©ν°λͺ¨λ¬ LLM νλ©΄ μΆλ‘ μμ΄μ νΈ, μ€ν 보μ μ‘°μκΈ° |
| Class 3: 리νλ μ΄ κ³΅κ²© (Replay) | Threat Flow | 99.95% μ°¨λ¨μ¨ | FNR: 0.05% |
μ΄μ μΈμ κΆ€μ 볡μ μ£Όμ , ν¨ν· νμ΄λ° μ‘°μ μ€νΈν |
β‘ μ’ ν© λ°©μ΄ μ§ν μμ½
- μ€μ μ’
ν© λ΄ μ°¨λ¨μ¨ (Attack Block Rate): 99.94% (λ―Ένλ₯ FNR:
0.06%) - μ μ μ¬μ©μ ν΅κ³Όμ¨ (Benign Pass Rate): 99.82% (μ€νλ₯ FPR:
0.18%) - Vision-LLM μμ΄μ νΈ μ°¨λ¨μ¨: 99.88%
- μ΅μ’ κ²μ¦ λ³΅ν© μ μ (Overall Score): 99.88 / 100.00
β‘ Inference Speed & Runtime Benchmark (μΆλ‘ λ²€μΉλ§ν¬)
| Runtime Engine | Single Inference Latency | Model Binary Size | Runtime Characteristics |
|---|---|---|---|
| CDN Dynamic INT8 ONNX | 0.42 ms | ~320 KiB |
Edge Worker(Cloudflare/Fastly), WASM λΈλΌμ°μ μΈλΌμΈ λ°°ν¬ μ΅μ ν |
| Standard ONNX (CPU) | 0.85 ms | ~1.25 MiB |
Node.js, Python, Go API κ²μ΄νΈμ¨μ΄ λ°±μλ νΈν |
| PyTorch SafeTensors | 1.10 ms | ~1.28 MiB |
κ³ μ λ‘λ© λ° κ°μ€μΉ 보μ κ²μ¦ λ€μ΄ν°λΈ ν μ |
μ΄μ μ§μ° CDN μ£μ§ ꡬλ: 320KB μμ€μ μ΄κ²½λ INT8 ONNX λ°μ΄λ리λ₯Ό ν΅ν΄ λΈλΌμ°μ μΈμ κ²μ¦ λλ νλ‘ νΈμλ CDN μ£μ§ λ¨μμ 0.5ms μ΄λ΄λ‘ μ¦κ°μ μΈ νλ³μ μνν©λλ€.
π¦ Export Formats & Deployment Artifacts (λ°°ν¬ νμΌ κ΅¬μ±)
model.safetensors: κ³ μ λ‘λ© λ° λ³΄μ κ²μ¦μ ν΅κ³Όν μμ κ°μ€μΉ ν μ νμΌmodel.onnx: ν¬λ‘μ€ νλ«νΌ νμ€ ONNX λͺ¨λΈ κ·Έλνmodel_cdn_int8.onnx: CDN Edge Worker λ° WASM νκ²½μ© INT8 μ΄κ²½λ λ°μ΄λ리 (~320 KiB)config.json: μ μΆλ ₯ κ·κ²© λ° ν΄λμ€ λ μ΄λΈ λ§€ν λ©νλ°μ΄ν°
π License & Commercial Terms (λΌμ΄μ μ€ λ° μμ μ λμ μλ΄)
Non-Commercial Use Only (λΉμμ μ μ°κ΅¬Β·κ°μΈ μ΄μ© νμ ):
This model is strictly provided for non-commercial, research, and educational purposes only. Any commercial use without a separate commercial agreement is strictly prohibited.
λ³Έ λͺ¨λΈμ μμ μ°κ΅¬, κ΅μ‘ λ° κ°μΈ λΉμ리 λͺ©μ μ νν΄μλ§ λ¬΄λ£λ‘ μ 곡λ©λλ€. μ¬μ νκ° μλ μΌμ²΄μ μ리μ μ΄μ©μ μ격ν κΈμ§λ©λλ€.Commercial Licensing & Inquiries for Revenue-Generating Entities (λ§€μΆ λ°μ κΈ°μ λ° μμ©ν λμ λ¬Έμ):
If your organization generates revenue, or if you plan to incorporate this model into commercial products, services, or paid APIs, you are required to purchase a commercial license. Please contact us directly via email for enterprise licensing and integration support.
κΈ°μ /μ¬μ 체μ λ§€μΆμ΄ λ°μνκ³ μκ±°λ, λ³Έ λͺ¨λΈμ μμ© μλΉμ€Β·μ λ£ μ νΒ·μμ΅ μ°½μΆ λͺ©μ μ μμ€ν μ λμ νκ³ μ νμλ κ²½μ° λ°λμ λ³λμ μμ© λΌμ΄μ μ€λ₯Ό μ·¨λνμ μΌ ν©λλ€. λμ 쑰건 λ° κΈ°μ μ© λΌμ΄μ μ€ λ°κΈμ μλ μ΄λ©μΌλ‘ λ¬Έμν΄ μ£ΌμκΈ° λ°λλλ€.- λμ λ° μμ© λΌμ΄μ μ€ λ¬Έμ (Commercial Inquiries): admin@099.kr
- 곡μ μΉμ¬μ΄νΈ (Official Website): https://baar.uk/model/
β οΈ Notes & Disclaimer (κΈ°μ κ³ μ§ λ° μλ΄)
- Defensive Engineering & Evaluation: λ³Έ λͺ¨λΈμ μ λ‘ νΈλ¬μ€νΈ λ΄ μ°¨λ¨ λ° μ΄μ νμ νμ§ μ°κ΅¬λ₯Ό μν΄ μ μλμμ΅λλ€. μ€μ μ΄μ νκ²½ λ°°μΉ μ, μμ¬ μλΉμ€ νΈλν½ νΉμ±μ λ§μΆμ΄ μ°¨λ¨ μκ³κ°(
config.json)μ μ¬μ μ μ‘°μ νμ¬ μ μ©νμμμ€.
πΌ Opportunities & Contact (μ±μ© μ μ λ° ννΈλμ)
This project demonstrates end-to-end competency in multimodal behavioral threat defense, bot heuristic spoofing mitigation, edge quantization (INT8/WASM), and ultra-low latency deployment.
I am actively seeking AI Engineering / Security AI Research opportunities, recruitment offers, and technical partnerships.
- 곡μ μΉμ¬μ΄νΈ (Official Website): https://baar.uk/model/
- κΈ°μ λμ λ° λΌμ΄μ μ€ λ¬Έμ (Commercial Adoption): admin@099.kr
- μ±μ© λ° ν¬μ§μ μ μ (Recruitment & Hiring): admin@099.kr
- ν¬μ λ° κΈ°μ νλ ₯ λ¬Έμ (Investment & Partnerships): admin@099.kr
κ·Ήλ¨μ μΈ λ¦¬μμ€ νκ²½(Edge, CDN, WASM)μμλ μ€μκ° μΆλ‘ μ΄ κ°λ₯ν λͺ¨λΈμ μ€κ³νκ³ , μ μ’ λ΄/μμ΄μ νΈ μν λ°©μ΄ μμ§ κ΅¬μΆλΆν° INT8 μμν λ° κΈλ‘λ² λ°°ν¬κΉμ§ μ κ³Όμ μ μ§μ ꡬννλ AI μμ§λμ΄μ λλ€. κΈ°μ λμ , μ±μ© μ μ λ° νμ μ κ΄μ¬ μλ κΈ°μ κ³Ό νμ λ¬Έμλ₯Ό νμν©λλ€.
- Downloads last month
- 23