Cluster and Aggregate: Face Recognition with Large Probe Set

19 Oct 2022  ยท  Minchul Kim, Feng Liu, Anil Jain, Xiaoming Liu ยท

Feature fusion plays a crucial role in unconstrained face recognition where inputs (probes) comprise of a set of $N$ low quality images whose individual qualities vary. Advances in attention and recurrent modules have led to feature fusion that can model the relationship among the images in the input set. However, attention mechanisms cannot scale to large $N$ due to their quadratic complexity and recurrent modules suffer from input order sensitivity. We propose a two-stage feature fusion paradigm, Cluster and Aggregate, that can both scale to large $N$ and maintain the ability to perform sequential inference with order invariance. Specifically, Cluster stage is a linear assignment of $N$ inputs to $M$ global cluster centers, and Aggregation stage is a fusion over $M$ clustered features. The clustered features play an integral role when the inputs are sequential as they can serve as a summarization of past features. By leveraging the order-invariance of incremental averaging operation, we design an update rule that achieves batch-order invariance, which guarantees that the contributions of early image in the sequence do not diminish as time steps increase. Experiments on IJB-B and IJB-S benchmark datasets show the superiority of the proposed two-stage paradigm in unconstrained face recognition. Code and pretrained models are available in https://github.com/mk-minchul/caface

PDF Abstract

Results from the Paper


 Ranked #1 on Face Verification on IJB-B (TAR @ FAR=0.001 metric)

     Get a GitHub badge
Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Face Verification IJB-B CAFace+AdaFace (WebFace4M) TAR @ FAR=0.001 96.91 # 1
TAR@FAR=0.0001 95.53 # 5
TAR @ FAR=1e-5 92.29 # 1
Face Verification IJB-C CAFace+AdaFace (WebFace4M) TAR @ FAR=1e-3 98.08 # 1
TAR @ FAR=1e-4 97.3% # 6
Surveillance-to-Surveillance IJB-S CAFace (WebFace4M) Rank-1 36.51 # 1
Rank-5 49.59 # 1
TAR @ FAR=0.01 8.78 # 1

Methods


No methods listed for this paper. Add relevant methods here