01
Send any image
Logo, label, packaging, product photo, or sketch. Vision® accepts any format and automatically extracts the relevant visual elements.
TrademarkIQ Vision®
Visual Recognition
of Trademarks by Artificial Intelligence.
Upload any image — logo, label, packaging, or sketch. The AI compares it against 30M+ trademarks indexed in Brazil and worldwide and identifies visual conflicts that the human eye cannot detect.
AI trained on 30M+ trademarks — high-dimensional embeddings for precise visual conflict detection.
30M+
indexed trademarks
Brazil and 60+ jurisdictions
3.1M+
figurative trademarks
with visual embedding
768
dimensions per image
high-fidelity TIQ-V2
<2s
response time
real-time visual search
How It Works
From Image to Insight in Seconds.
Four automated stages — from upload to strategic decision.
02
Comparison against 30M+ trademarks
768-dimensional TIQ-V2 embeddings compare your image against the entire indexed figurative repository — Brazil and 60+ jurisdictions in real time.
03
Visual conflicts identified
The system returns visually similar trademarks with a proximity score, Nice class, status, and jurisdiction. Conflicts that the human eye does not detect.
04
Complete strategic context
Each result includes the owner, attorney, dispatch history, and opposition window — everything you need to decide before acting.
Capabilities
Any trademark. Any image.
Anywhere in the world.
From logos to industrial designs — Vision® recognizes protected visual elements across the entire spectrum of intellectual property.
Logos and figurative trademarks
Detection of similarity in graphic elements, stylized typography, icons, and registered visual compositions.
Industrial designs
Recognition of shapes, ornamental patterns, and visual configurations protected as industrial designs in the INPI and international offices.
Packaging and trade dress
Image set analysis: dominant colors, element arrangement, and overall product visual identity.
Global coverage
Visual search in 60+ jurisdictions. Find conflicts before expanding into new markets — from the EUIPO to the WIPO, from the USPTO to the INPI.
Use cases
Visual intelligence for every step of trademark protection.
Visual prior art search
Before filing, check if there is a visually similar trademark already registered. Reduce the risk of rejection and opposition with concrete visual evidence.
Reduce risk before filing
Portfolio monitoring
Monitor new publications from the INPI and international offices. Receive alerts when a visually similar trademark to your portfolio is filed.
Real-time alerts
Competitive intelligence
Analyze the visual movements of competitors. Identify branding trends in your segment and anticipate positioning strategies.
Real competitive advantage
Technology
High-fidelity visual embeddings.
Each indexed trademark is transformed into a 768-dimensional vector by the TIQ-V2 model — capturing shape, composition, typography, and graphic elements with precision superior to any textual search.
- TIQ-V2 — TrademarkIQ Vision 2High-capacity model trained to capture visual nuances in registered trademarks.
- Cosine similarity searchAccurate vector distance — orders results by actual degree of visual similarity.
- Optimized IVFFlat indexingSub-second search over millions of embeddings with pgvector and specialized indices.
Model
TIQ-V2
TrademarkIQ Vision 2
Dimensions
768
halfvec — high accuracy
Figurative marks indexed
3,1M+
Complete figurative repository
Response time
<2s
Real-time visual search
TrademarkIQ Vision®
See what no one else sees. Protect before it's too late.
Upload your first image and discover visual conflicts in seconds. Available in all TrademarkIQ plans.