[Example] Human skeletal authentication
The required period is 3 months! We will introduce a case where a model was trained using a dataset!
As a challenge, it was pointed out that "we want to use AI (OpenPose) for human skeletal authentication, but using the existing fisheye lens (full-surround lens) camera significantly reduces the accuracy of the AI (skeletal authentication)." In response, our company considered the development policy and created a training dataset. After training the model using the dataset, we conducted learning evaluation, judgment result evaluation, and tuning. 【Case Overview】 ■Challenge - We want to use AI for human skeletal authentication, but using the existing fisheye lens camera significantly reduces the accuracy of skeletal authentication. - Images taken with a full-surround lens can only recognize skeletal features within a 45° range directly above. ■Tasks - Consideration of development policy - Creation of training dataset - Model training, evaluation, and tuning *For more details, please refer to the PDF document or feel free to contact us.
- Company:テクノプロ テクノプロ・デザイン社【東証プライム上場グループ】
- Price:Other