Cases

There is a high requirement for the functionality and convenience of the labeling tool. The customer handed us an urgent task during a holiday season, the customers needed a large amount of special scenario data urgently.

Based on the full excavation and sorting of customer needs, according to the scenario classification, we provided the optimal service program to cooperate with the customer's algorithmic products for rapid delivery, and delivers the data with high quality within 4 days.

Within 24 hours of the client's unexpected request, we responded quickly to provide a project scheduling plan and implemented all of it, leading to the successful delivery of the client's algorithmic products.

Exploratory projects are take up a large portion and require mostly specialized terminology that needs to be translated into common parlance and project rules that are consistent with algorithmic requirements.

Reorganize the project rules according to the algorithm requirements, convert the specialized terminology into common vocabulary and describe it with sample diagrams, which will be confirmed by the client algorithm.

Help clients form sample rule templates, reduce communication costs, execute projects efficiently, and deliver data that satisfies clients.

More than 20 requirements are presented, and the labeling rules are accordingly subdivided into more than 20 categories, providing greater difficulty in project execution.

Complete mastery of the rules, combining more than 20 subdivisions into 5 major programs and integrating 5 uniform rules.

Avoid the possibility of confusion caused by multiple projects, significantly improve the efficiency of project execution, and shorten the algorithm development time for clients.

Rules are affected by multiple scenarios, attributes are judged more autonomously, and we have to complete analysis of scenarios that affect the effectiveness of the data for the client.

The rules are categorized according to the scenarios, and the annotators are trained and practiced point by point to finally reach full mastery of the rules. In addition, real-time scene data information analysis report is provided.

Assist algorithms to do scenario problem localization, and help clients' algorithm development process to move forward smoothly.

The correctness statistics logic of the customer's own platform is different from that of ordinary platforms in the market, and the correctness requirement of 99% requires higher comprehensive quality of labelers and auditors.

Through personnel training, personnel screening, and a proven audit and quality assurance fission program, we are able to achieve the best of the best in terms of annotators and auditors, and meet the project's operational requirements.

We meet the client's project requirements, deliver with high efficiency and quality, and further accumulate our experience in adopting different clients' own platforms.

Infrared camera data, rules and regular requirements differ more, and subjective judgment is greater. Requirements for personnel subjectivity are greater.

Sort out and categorize data scenarios, summarize segments involving subjective judgments into sample sets, confirm with customers point by point, and provide labeling training support.

The client highly appreciated the project delivery results and the company accumulated valuable experience in infrared data labeling.