MiVOLO: Multi-input Transformer for Age and Gender Estimation

10 Jul 2023  ·  Maksim Kuprashevich, Irina Tolstykh ·

Age and gender recognition in the wild is a highly challenging task: apart from the variability of conditions, pose complexities, and varying image quality, there are cases where the face is partially or completely occluded. We present MiVOLO (Multi Input VOLO), a straightforward approach for age and gender estimation using the latest vision transformer. Our method integrates both tasks into a unified dual input/output model, leveraging not only facial information but also person image data. This improves the generalization ability of our model and enables it to deliver satisfactory results even when the face is not visible in the image. To evaluate our proposed model, we conduct experiments on four popular benchmarks and achieve state-of-the-art performance, while demonstrating real-time processing capabilities. Additionally, we introduce a novel benchmark based on images from the Open Images Dataset. The ground truth annotations for this benchmark have been meticulously generated by human annotators, resulting in high accuracy answers due to the smart aggregation of votes. Furthermore, we compare our model's age recognition performance with human-level accuracy and demonstrate that it significantly outperforms humans across a majority of age ranges. Finally, we grant public access to our models, along with the code for validation and inference. In addition, we provide extra annotations for used datasets and introduce our new benchmark.

PDF Abstract


Introduced in the Paper:


Used in the Paper:

UTKFace FairFace Adience AgeDB IMDB-Clean

Results from the Paper

 Ranked #1 on Age Estimation on AgeDB (using extra training data)

     Get a GitHub badge
Task Dataset Model Metric Name Metric Value Global Rank Uses Extra
Training Data
Result Benchmark
Age And Gender Classification Adience Age MiVOLO-D1 Accuracy (5-fold) 68.69 # 1
Age And Gender Classification Adience Gender MiVOLO-D1 Accuracy (5-fold) 96.51 # 1
Age Estimation AgeDB MiVOLO-D1 MAE 5.55 # 1
Gender Prediction AgeDB MiVOLO-D1 Accuracy 98.3 # 1
Facial Attribute Classification FairFace MiVOLO-D1 gender-top1 95.73 # 1
age-top1 61.07 # 1
Age Estimation IMDB-Clean VOLO-D1 age&gender Average mean absolute error 4.22 # 2
Age Estimation IMDB-Clean MiVOLO-D1 Average mean absolute error 4.09 # 1
Age Estimation LAGENDA MiVOLO-D1 MAE 3.99 # 1
Gender Prediction LAGENDA MiVOLO-D1 Accuracy 97.36 # 1
Age and Gender Estimation LAGENDA age MiVOLO-D1 MAE 3.99 # 1
CS@5 71.27 # 1
Age and Gender Estimation LAGENDA gender MiVOLO-D1 Accuracy 97.36 # 1
Age Estimation UTKFace VOLO-D1 age&gender MAE 4.23 # 3
Age Estimation UTKFace MiVOLO-D1 MAE 3.7 # 1