Cadence® system design and verification solutions, integrated under our Verification Suite, provide the simulation, acceleration, emulation, and management capabilities.
Verification Suite Related Products A-Z
Cadence® digital design and signoff solutions provide a fast path to design closure and better predictability, helping you meet your power, performance, and area (PPA) targets.
Full-Flow Digital Solution Related Products A-Z
Cadence® custom, analog, and RF design solutions can help you save time by automating many routine tasks, from block-level and mixed-signal simulation to routing and library characterization.
Overview Related Products A-Z
Driving efficiency and accuracy in advanced packaging, system planning, and multi-fabric interoperability, Cadence® package implementation products deliver the automation and accuracy.
Cadence® PCB design solutions enable shorter, more predictable design cycles with greater integration of component design and system-level simulation for a constraint-driven flow.
An open IP platform for you to customize your app-driven SoC design.
Comprehensive solutions and methodologies.
Helping you meet your broader business goals.
A global customer support infrastructure with around-the-clock help.
More Support Log In
24/7 Support - Cadence Online Support
Locate the latest software updates, service request, technical documentation, solutions and more in your personalized environment.
Cadence offers various software services for download. This page describes our offerings, including the Allegro FREE Physical Viewer.
The Cadence Academic Network helps build strong relationships between academia and industry, and promotes the proliferation of leading-edge technologies and methodologies at universities renowned for their engineering and design excellence.
Participate in CDNLive
A huge knowledge exchange platform for academia to network with industry. We are looking for academic speakers to talk about their research to the industry attendees at the Academic Track at CDNLive EMEA and Silicon Valley.
Come & Meet Us @ Events
A huge knowledge exchange platform for academia. We are looking for academic speakers to talk about their research to industry attendees.
Americas University Software Program
Join the 250+ qualified Americas member universities who have already incorporated Cadence EDA software into their classrooms and academic research projects.
EMEA University Software Program
In EMEA, Cadence works with EUROPRACTICE to ensure cost-effective availability of our extensive electronic design automation (EDA) tools for non-commercial activities.
Apply Now For Jobs
If you are a recent college graduate or a student looking for internship. Visit our exclusive job search page for interns and recent college graduate jobs.
Cadence is a Great Place to do great work
Learn more about our internship program and visit our careers page to do meaningful work and make a great impact.
Get the most out of your investment in Cadence technologies through a wide range of training offerings.
Overview All Courses Asia Pacific EMEANorth America
Instructor-led training [ILT] are live classes that are offered in our state-of-the-art classrooms at our worldwide training centers, at your site, or as a Virtual classroom.
Online Training is delivered over the web to let you proceed at your own pace, anytime and anywhere.
Exchange ideas, news, technical information, and best practices.
The community is open to everyone, and to provide the most value, we require participants to follow our Community Guidelines that facilitate a quality exchange of ideas and information.
It's not all about the technology. Here we exchange ideas on the Cadence Academic Network and other subjects of general interest.
Cadence is a leading provider of system design tools, software, IP, and services.
Get email delivery of the Cadence blog featured here
The German Traffic Sign Benchmark actually has the signs divided into groups: speed limits, danger signs, and so on. It turns out that humans make far fewer errors between groups than within groups. They might mistake a 30kph speed limit sign for an 80kph sign, but very rarely for a stop sign. The errors CNNs make are much less structured and are pretty much all over the place, so there is clearly room for improvement. How can we teach CNNs to be more like humans?
The approach is to build a hierarchical CNN using a divide-and-conquer approach. One CNN is used to classify the sign into a group, and then each group has its own CNN for classification within the group. It may sound like this is duplicating a lot of work and is likely to make things worse, but the resulting networks are a lot simpler (and, of course, can be optimized in the ways outlined in yesterday’s post).
It turns out that the families that humans use naturally are not the best for this approach, and a CNN family clustering (CFC) algorithm can be used to develop better families. Interestingly, this is something Netflix discovered. When they started, they classified movies as romance, comedy, action, etc. and tried to decide how much each person likes each of those classifications. When they had a competition to improve their recommendation system, the leaders all ignored that information and independently clustered movies into groups that seemed to have predictive power, but didn’t necessarily make sense to humans.
A simplified view of the algorithm is:
So on the street signs you can see the first split into a seed set A and a seed set B. Seed set B, once extended with additional members, is a “good” family, but there is too much variation in seed set A and it needs to be further split.
Once the process is complete, we end up with five families. Some of the clustering, such as all the triangular warning signs, are similar to what a human would create. But there are some unexpected groupings. For example, seed set B in the above diagram is signs with a diagonal stripe from top right to bottom left, and ends up including the keep left sign that a human would most likely group with the other directional signs.
The training and recognition phases are fairly obvious. The family classification CNN is trained using all the training data. Then a CNN for each family is trained using just the data for that family. During inference, the family classification CNN is used to identify the family and then the appropriate CNN is used to complete recognition. See the picture above.
One reason that this process is effective is that after the hierarchical decomposition into families, more resources can be allocated in relation to the problem difficulty to the per family CNNs and this is more effective than just having a huge CNN for the entire set of signs, or a sub-optimal classification into families that are convenient for people, but which may leave a few signs in each family that are especially hard to identify.
The results are, indeed, improved. Using the pre-defined (human) families, the misclassification across families was 0.23% and within families was 0.33%. With the CFC created families (in the pictures above) the error rates are reduced to 0.1% and 0.24%, respectively.
It turns out that this Cadence HCNN is the current leader (probably not for long) in recognition for the German street sign benchmark, achieving a recognition rate of 99.82% with a complexity of only 83 million MACS. Traffic sign recognition neural networks exceed human performance.
Next: The Future of Neural Networks...and Our Robot Overlords
Previous: How to Optimize Your CNN