catastrophic forgetting
multi-task
- computation problem ---- using all data
- storage issue ---- always keep the data
- multi-task traning can be considered as the upper bound of LLL
multi-model
- storage issue ---- can not store all models
- Knowledge cannot transfer across different models
Transfer vs LLL
- learn task2 based on the task1
- learn task2 not forgetting task1
Evaluation
![在这里插入图片描述](https://img-blog.csdnimg.cn/d2377a2568434a908201e465b08c0b98.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzQwNDM4NTIz,size_16,color_FFFFFF,t_70#pic_center)
Selective Synaptic Plasticity
- Basic idea: only change the unimportant parameters
- method: update the loss function
L
′
(
θ
)
=
L
(
θ
)
+
λ
∑
i
b
i
(
θ
i
?
θ
i
b
)
2
L'(\theta)=L(\theta)+\lambda \sum_{i}b_i(\theta_{i}-\theta_i^b)^2
L′(θ)=L(θ)+λi∑?bi?(θi??θib?)2
θ
i
b
\theta_i^b
θib? is the model learned from the previous tasks - Gradient Episodic Memory(GEM)
![在这里插入图片描述](https://img-blog.csdnimg.cn/ad51c2a9ee404dc0a4577ec66693d2bf.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzQwNDM4NTIz,size_16,color_FFFFFF,t_70#pic_center)
Additional Neural Resource Allocation
-
Progressive Neural Networks ![在这里插入图片描述](https://img-blog.csdnimg.cn/90b01e3d86cc4136aa3c410dc93720e4.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzQwNDM4NTIz,size_16,color_FFFFFF,t_70) Add new parameters based on previous model -
PackNet From the begining, we set a large model with enough parameters and update different parameters every task. ![在这里插入图片描述](https://img-blog.csdnimg.cn/82b21fcd8bf04aa69822b2600bd639bf.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzQwNDM4NTIz,size_16,color_FFFFFF,t_70)
Memory Reply
- generating data:
We need a additional generator to generate data for the task before ![在这里插入图片描述](https://img-blog.csdnimg.cn/89e8efdb4798464fb788022a2d846f13.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzQwNDM4NTIz,size_16,color_FFFFFF,t_70) - Adding new classes
|