All experiments are run on a cluster of 30 machines equipped with single RTX3070Ti GPUs.
All experiments are run on a cluster of 30 machines equipped with single RTX3070Ti GPUs.
% General experimental setup -> ML domain
% General experimental setup -> ML domain
Replay is investigated in a supervised CIL-scenario, assuming known task-boundaries and disjoint classes.
Replay is investigated in a supervised CIL-scenario, assuming known task-boundaries and disjoint classes. All of the following details apply to all investigated CL algorithms, namely AR, ER and DGR with VAEs.
% Balancing of Tasks/Classes
% Balancing of Tasks/Classes
Tasks $T_{i}$ contain all samples of the corresponding classes defining them, see \cref{tab:slts} for details.
Tasks $T_{i}$ contain all samples of the corresponding classes defining them, see \cref{tab:slts} for details.
% TODO: OK ???
It is assumed that data from all tasks occur with equal probability. Some datasets are slightly unbalanced, for example Fruits and SVHN classes 1 and 2, which may render certain sub-task settings as more difficult.
It is assumed that data from all tasks occurs with equal probability, however, it is not ensured that the amount/variability of samples per class is balanced, see e.g., SVHN classes 1 \& 2, which may render certain sub-task settings as more difficult.
% Initial/Replay
% Initial/Replay
Training consists of an (initial) run on $T_1$, followed by a sequence of independent (replay) runs on $T_{i>1}$.
Training consists of an (initial) run on $T_1$, followed by a sequence of independent (replay) runs on $T_{i>1}$.
% Averaged over runs & baseline experiments
% Averaged over runs & baseline experiments
...
@@ -275,7 +274,7 @@
...
@@ -275,7 +274,7 @@
It is worth noting that classes will, in general, \textit{not} be balanced in the merged generated/real data at $T_i$, and that it is not required to store the statictics of previously encountered class instances/labels.
It is worth noting that classes will, in general, \textit{not} be balanced in the merged generated/real data at $T_i$, and that it is not required to store the statictics of previously encountered class instances/labels.
\caption{\label{fig:vargen} An example for variant generation in AR, see \cref{sec:approach} and \cref{fig:var} for details. Left: centroids of the current GMM scholar trained on MNIST classes 0, 4 and 6. Middle: query samples of MNIST class 9. Right: variants generated in response to the query. Component weights and variances are not shown.
\caption{\label{fig:vargen} An example for variant generation in AR, see \cref{sec:approach} and \cref{fig:var} for details. Left: centroids of the current GMM scholar trained on MNIST classes 0, 4 and 6. Middle: query samples of MNIST class 9. Right: variants generated in response to the query. Component weights and variances are not shown.
}
}
\end{figure}
\end{figure}
First, we demonstrate the ability of a GMM layer $L_{(G)}$ to query its internal representation through data samples and selectively generate artificial data that \enquote{best match} those that define the query. To illustrate this, we train a GMM layer of $K=25$ components on MNIST classes 0,4 and 6 for 50 epochs using the best-practice rules described in \cref{app:ar}. Then, we query the trained GMM with samples from class 9 uniquely, as described in \cref{sec:gmm}. The resulting samples are all from class 4, since it is the class that is \enquote{most similar} to the query class. These results are visualized in \cref{fig:var}. Variant generation results for deep convolutional extensions of GMMs can be found in \cite{gepperth2021new}, emphasizing that the AR approach can be scaled to more complex problems.
First, we demonstrate the ability of a trained GMM to query its internal representation through data samples and selectively generate artificial data that \enquote{best match} those that define the query. To illustrate this, we train a GMM layer of $K=25$ components on MNIST classes 0,4 and 6 for 50 epochs using the best-practice rules described in \cref{app:ar}. Then, we query the trained GMM with samples from class 9 uniquely, as described in \cref{sec:gmm}. The resulting samples are all from class 4, since it is the class that is \enquote{most similar} to the query class. These results are visualized in \cref{fig:var}. Variant generation results for deep convolutional extensions of GMMs can be found in \cite{gepperth2021new}, emphasizing that the AR approach can be scaled to more complex problems.
