Classifying bodily exercise from smartphone knowledge

Introduction

On this put up we’ll describe the right way to use smartphone accelerometer and gyroscope knowledge to foretell the bodily actions of the people carrying the telephones. The information used on this put up comes from the Smartphone-Based mostly Recognition of Human Actions and Postural Transitions Information Set distributed by the College of California, Irvine. Thirty people had been tasked with performing varied primary actions with an connected smartphone recording motion utilizing an accelerometer and gyroscope.

Earlier than we start, let’s load the varied libraries that we’ll use within the evaluation:

library(keras)     # Neural Networks
library(tidyverse) # Information cleansing / Visualization
library(knitr)     # Desk printing
library(rmarkdown) # Misc. output utilities 
library(ggridges)  # Visualization

Actions dataset

The information used on this put up come from the Smartphone-Based mostly Recognition of Human Actions and Postural Transitions Information Set(Reyes-Ortiz et al. 2016) distributed by the College of California, Irvine.

When downloaded from the hyperlink above, the information comprises two totally different ‘components.’ One which has been pre-processed utilizing varied characteristic extraction strategies comparable to fast-fourier rework, and one other RawData part that merely provides the uncooked X,Y,Z instructions of an accelerometer and gyroscope. None of the usual noise filtering or characteristic extraction utilized in accelerometer knowledge has been utilized. That is the information set we’ll use.

The motivation for working with the uncooked knowledge on this put up is to assist the transition of the code/ideas to time sequence knowledge in much less well-characterized domains. Whereas a extra correct mannequin may very well be made by using the filtered/cleaned knowledge offered, the filtering and transformation can range significantly from activity to activity; requiring plenty of handbook effort and area data. One of many lovely issues about deep studying is the characteristic extraction is realized from the information, not exterior data.

Exercise labels

The information has integer encodings for the actions which, whereas not necessary to the mannequin itself, are useful to be used to see. Let’s load them first.

activityLabels <- learn.desk("knowledge/activity_labels.txt", 
                             col.names = c("quantity", "label")) 

activityLabels %>% kable(align = c("c", "l"))

Subsequent, we load within the labels key for the RawData. This file is a listing of all the observations, or particular person exercise recordings, contained within the knowledge set. The important thing for the columns is taken from the information README.txt.

Column 1: experiment quantity ID, 
Column 2: person quantity ID, 
Column 3: exercise quantity ID 
Column 4: Label begin level 
Column 5: Label finish level 

The beginning and finish factors are in variety of sign log samples (recorded at 50hz).

Let’s check out the primary 50 rows:

labels <- learn.desk(
  "knowledge/RawData/labels.txt",
  col.names = c("experiment", "userId", "exercise", "startPos", "endPos")
)

labels %>% 
  head(50) %>% 
  paged_table()

File names

Subsequent, let’s have a look at the precise information of the person knowledge offered to us in RawData/

dataFiles <- listing.information("knowledge/RawData")
dataFiles %>% head()

[1] "acc_exp01_user01.txt" "acc_exp02_user01.txt"
[3] "acc_exp03_user02.txt" "acc_exp04_user02.txt"
[5] "acc_exp05_user03.txt" "acc_exp06_user03.txt"

There’s a three-part file naming scheme. The primary half is the kind of knowledge the file comprises: both acc for accelerometer or gyro for gyroscope. Subsequent is the experiment quantity, and final is the person Id for the recording. Let’s load these right into a dataframe for ease of use later.

fileInfo <- data_frame(
  filePath = dataFiles
) %>%
  filter(filePath != "labels.txt") %>% 
  separate(filePath, sep = '_', 
           into = c("kind", "experiment", "userId"), 
           take away = FALSE) %>% 
  mutate(
    experiment = str_remove(experiment, "exp"),
    userId = str_remove_all(userId, "person|.txt")
  ) %>% 
  unfold(kind, filePath)

fileInfo %>% head() %>% kable()

Studying and gathering knowledge

Earlier than we will do something with the information offered we have to get it right into a model-friendly format. This implies we wish to have a listing of observations, their class (or exercise label), and the information comparable to the recording.

To acquire this we’ll scan via every of the recording information current in dataFiles, search for what observations are contained within the recording, extract these recordings and return all the pieces to a simple to mannequin with dataframe.

# Learn contents of single file to a dataframe with accelerometer and gyro knowledge.
readInData <- operate(experiment, userId){
  genFilePath = operate(kind) {
    paste0("knowledge/RawData/", kind, "_exp",experiment, "_user", userId, ".txt")
  }  
  
  bind_cols(
    learn.desk(genFilePath("acc"), col.names = c("a_x", "a_y", "a_z")),
    learn.desk(genFilePath("gyro"), col.names = c("g_x", "g_y", "g_z"))
  )
}

# Perform to learn a given file and get the observations contained alongside
# with their lessons.

loadFileData <- operate(curExperiment, curUserId) {
  
  # load sensor knowledge from file into dataframe
  allData <- readInData(curExperiment, curUserId)

  extractObservation <- operate(startPos, endPos){
    allData[startPos:endPos,]
  }
  
  # get statement areas on this file from labels dataframe
  dataLabels <- labels %>% 
    filter(userId == as.integer(curUserId), 
           experiment == as.integer(curExperiment))
  

  # extract observations as dataframes and save as a column in dataframe.
  dataLabels %>% 
    mutate(
      knowledge = map2(startPos, endPos, extractObservation)
    ) %>% 
    choose(-startPos, -endPos)
}

# scan via all experiment and userId combos and collect knowledge right into a dataframe. 
allObservations <- map2_df(fileInfo$experiment, fileInfo$userId, loadFileData) %>% 
  right_join(activityLabels, by = c("exercise" = "quantity")) %>% 
  rename(activityName = label)

# cache work. 
write_rds(allObservations, "allObservations.rds")
allObservations %>% dim()

Exploring the information

Now that now we have all the information loaded together with the experiment, userId, and exercise labels, we will discover the information set.

Size of recordings

Let’s first have a look at the size of the recordings by exercise.

allObservations %>% 
  mutate(recording_length = map_int(knowledge,nrow)) %>% 
  ggplot(aes(x = recording_length, y = activityName)) +
  geom_density_ridges(alpha = 0.8)

The very fact there’s such a distinction in size of recording between the totally different exercise sorts requires us to be a bit cautious with how we proceed. If we prepare the mannequin on each class directly we’re going to must pad all of the observations to the size of the longest, which would go away a big majority of the observations with an enormous proportion of their knowledge being simply padding-zeros. Due to this, we’ll match our mannequin to simply the biggest ‘group’ of observations size actions, these embody STAND_TO_SIT, STAND_TO_LIE, SIT_TO_STAND, SIT_TO_LIE, LIE_TO_STAND, and LIE_TO_SIT.

An fascinating future path could be trying to make use of one other structure comparable to an RNN that may deal with variable size inputs and coaching it on all the information. Nonetheless, you’ll run the chance of the mannequin studying merely that if the statement is lengthy it’s almost certainly one of many 4 longest lessons which might not generalize to a situation the place you had been operating this mannequin on a real-time-stream of knowledge.

Filtering actions

Based mostly on our work from above, let’s subset the information to simply be of the actions of curiosity.

desiredActivities <- c(
  "STAND_TO_SIT", "SIT_TO_STAND", "SIT_TO_LIE", 
  "LIE_TO_SIT", "STAND_TO_LIE", "LIE_TO_STAND"  
)

filteredObservations <- allObservations %>% 
  filter(activityName %in% desiredActivities) %>% 
  mutate(observationId = 1:n())

filteredObservations %>% paged_table()

So after our aggressive pruning of the information we could have a good quantity of knowledge left upon which our mannequin can be taught.

Coaching/testing cut up

Earlier than we go any additional into exploring the information for our mannequin, in an try and be as truthful as attainable with our efficiency measures, we have to cut up the information right into a prepare and check set. Since every person carried out all actions simply as soon as (excluding one who solely did 10 of the 12 actions) by splitting on userId we’ll be certain that our mannequin sees new individuals solely once we check it.

# get all customers
userIds <- allObservations$userId %>% distinctive()

# randomly select 24 (80% of 30 people) for coaching
set.seed(42) # seed for reproducibility
trainIds <- pattern(userIds, dimension = 24)

# set the remainder of the customers to the testing set
testIds <- setdiff(userIds,trainIds)

# filter knowledge. 
trainData <- filteredObservations %>% 
  filter(userId %in% trainIds)

testData <- filteredObservations %>% 
  filter(userId %in% testIds)

Visualizing actions

Now that now we have trimmed our knowledge by eradicating actions and splitting off a check set, we will truly visualize the information for every class to see if there’s any instantly discernible form that our mannequin could possibly decide up on.

First let’s unpack our knowledge from its dataframe of one-row-per-observation to a tidy model of all of the observations.

unpackedObs <- 1:nrow(trainData) %>% 
  map_df(operate(rowNum){
    dataRow <- trainData[rowNum, ]
    dataRow$knowledge[[1]] %>% 
      mutate(
        activityName = dataRow$activityName, 
        observationId = dataRow$observationId,
        time = 1:n() )
  }) %>% 
  collect(studying, worth, -time, -activityName, -observationId) %>% 
  separate(studying, into = c("kind", "path"), sep = "_") %>% 
  mutate(kind = ifelse(kind == "a", "acceleration", "gyro"))

Now now we have an unpacked set of our observations, let’s visualize them!

unpackedObs %>% 
  ggplot(aes(x = time, y = worth, coloration = path)) +
  geom_line(alpha = 0.2) +
  geom_smooth(se = FALSE, alpha = 0.7, dimension = 0.5) +
  facet_grid(kind ~ activityName, scales = "free_y") +
  theme_minimal() +
  theme( axis.textual content.x = element_blank() )

So a minimum of within the accelerometer knowledge patterns positively emerge. One would think about that the mannequin might have hassle with the variations between LIE_TO_SIT and LIE_TO_STAND as they’ve an identical profile on common. The identical goes for SIT_TO_STAND and STAND_TO_SIT.

Preprocessing

Earlier than we will prepare the neural community, we have to take a few steps to preprocess the information.

Padding observations

First we’ll determine what size to pad (and truncate) our sequences to by discovering what the 98th percentile size is. By not utilizing the very longest statement size it will assist us keep away from extra-long outlier recordings messing up the padding.

padSize <- trainData$knowledge %>% 
  map_int(nrow) %>% 
  quantile(p = 0.98) %>% 
  ceiling()
padSize

98% 
334 

Now we merely must convert our listing of observations to matrices, then use the tremendous useful pad_sequences() operate in Keras to pad all observations and switch them right into a 3D tensor for us.

convertToTensor <- . %>% 
  map(as.matrix) %>% 
  pad_sequences(maxlen = padSize)

trainObs <- trainData$knowledge %>% convertToTensor()
testObs <- testData$knowledge %>% convertToTensor()
  
dim(trainObs)

[1] 286 334   6

Fantastic, we now have our knowledge in a pleasant neural-network-friendly format of a 3D tensor with dimensions (<num obs>, <sequence size>, <channels>).

One-hot encoding

There’s one final thing we have to do earlier than we will prepare our mannequin, and that’s flip our statement lessons from integers into one-hot, or dummy encoded, vectors. Fortunately, once more Keras has provided us with a really useful operate to do exactly this.

oneHotClasses <- . %>% 
  {. - 7} %>%        # deliver integers all the way down to 0-6 from 7-12
  to_categorical() # One-hot encode

trainY <- trainData$exercise %>% oneHotClasses()
testY <- testData$exercise %>% oneHotClasses()

Modeling

Structure

Since now we have temporally dense time-series knowledge we’ll make use of 1D convolutional layers. With temporally-dense knowledge, an RNN has to be taught very lengthy dependencies with a view to decide up on patterns, CNNs can merely stack a number of convolutional layers to construct sample representations of considerable size. Since we’re additionally merely on the lookout for a single classification of exercise for every statement, we will simply use pooling to ‘summarize’ the CNNs view of the information right into a dense layer.

Along with stacking two layer_conv_1d() layers, we’ll use batch norm and dropout (the spatial variant(Tompson et al. 2014) on the convolutional layers and commonplace on the dense) to regularize the community.

input_shape <- dim(trainObs)[-1]
num_classes <- dim(trainY)[2]

filters <- 24     # variety of convolutional filters to be taught
kernel_size <- 8  # what number of time-steps every conv layer sees.
dense_size <- 48  # dimension of our penultimate dense layer. 

# Initialize mannequin
mannequin <- keras_model_sequential()
mannequin %>% 
  layer_conv_1d(
    filters = filters,
    kernel_size = kernel_size, 
    input_shape = input_shape,
    padding = "legitimate", 
    activation = "relu"
  ) %>%
  layer_batch_normalization() %>%
  layer_spatial_dropout_1d(0.15) %>% 
  layer_conv_1d(
    filters = filters/2,
    kernel_size = kernel_size,
    activation = "relu",
  ) %>%
  # Apply common pooling:
  layer_global_average_pooling_1d() %>% 
  layer_batch_normalization() %>%
  layer_dropout(0.2) %>% 
  layer_dense(
    dense_size,
    activation = "relu"
  ) %>% 
  layer_batch_normalization() %>%
  layer_dropout(0.25) %>% 
  layer_dense(
    num_classes, 
    activation = "softmax",
    identify = "dense_output"
  ) 

abstract(mannequin)

______________________________________________________________________
Layer (kind)                   Output Form                Param #    
======================================================================
conv1d_1 (Conv1D)              (None, 327, 24)             1176       
______________________________________________________________________
batch_normalization_1 (BatchNo (None, 327, 24)             96         
______________________________________________________________________
spatial_dropout1d_1 (SpatialDr (None, 327, 24)             0          
______________________________________________________________________
conv1d_2 (Conv1D)              (None, 320, 12)             2316       
______________________________________________________________________
global_average_pooling1d_1 (Gl (None, 12)                  0          
______________________________________________________________________
batch_normalization_2 (BatchNo (None, 12)                  48         
______________________________________________________________________
dropout_1 (Dropout)            (None, 12)                  0          
______________________________________________________________________
dense_1 (Dense)                (None, 48)                  624        
______________________________________________________________________
batch_normalization_3 (BatchNo (None, 48)                  192        
______________________________________________________________________
dropout_2 (Dropout)            (None, 48)                  0          
______________________________________________________________________
dense_output (Dense)           (None, 6)                   294        
======================================================================
Whole params: 4,746
Trainable params: 4,578
Non-trainable params: 168
______________________________________________________________________

Coaching

Now we will prepare the mannequin utilizing our check and coaching knowledge. Word that we use callback_model_checkpoint() to make sure that we save solely the most effective variation of the mannequin (fascinating since in some unspecified time in the future in coaching the mannequin might start to overfit or in any other case cease bettering).

# Compile mannequin
mannequin %>% compile(
  loss = "categorical_crossentropy",
  optimizer = "rmsprop",
  metrics = "accuracy"
)

trainHistory <- mannequin %>%
  match(
    x = trainObs, y = trainY,
    epochs = 350,
    validation_data = listing(testObs, testY),
    callbacks = listing(
      callback_model_checkpoint("best_model.h5", 
                                save_best_only = TRUE)
    )
  )

The mannequin is studying one thing! We get a good 94.4% accuracy on the validation knowledge, not dangerous with six attainable lessons to select from. Let’s look into the validation efficiency slightly deeper to see the place the mannequin is messing up.

Analysis

Now that now we have a educated mannequin let’s examine the errors that it made on our testing knowledge. We are able to load the most effective mannequin from coaching primarily based upon validation accuracy after which have a look at every statement, what the mannequin predicted, how excessive a chance it assigned, and the true exercise label.

# dataframe to get labels onto one-hot encoded prediction columns
oneHotToLabel <- activityLabels %>% 
  mutate(quantity = quantity - 7) %>% 
  filter(quantity >= 0) %>% 
  mutate(class = paste0("V",quantity + 1)) %>% 
  choose(-number)

# Load our greatest mannequin checkpoint
bestModel <- load_model_hdf5("best_model.h5")

tidyPredictionProbs <- bestModel %>% 
  predict(testObs) %>% 
  as_data_frame() %>% 
  mutate(obs = 1:n()) %>% 
  collect(class, prob, -obs) %>% 
  right_join(oneHotToLabel, by = "class")

predictionPerformance <- tidyPredictionProbs %>% 
  group_by(obs) %>% 
  summarise(
    highestProb = max(prob),
    predicted = label[prob == highestProb]
  ) %>% 
  mutate(
    fact = testData$activityName,
    appropriate = fact == predicted
  ) 

predictionPerformance %>% paged_table()

First, let’s have a look at how ‘assured’ the mannequin was by if the prediction was appropriate or not.

predictionPerformance %>% 
  mutate(consequence = ifelse(appropriate, 'Right', 'Incorrect')) %>% 
  ggplot(aes(highestProb)) +
  geom_histogram(binwidth = 0.01) +
  geom_rug(alpha = 0.5) +
  facet_grid(consequence~.) +
  ggtitle("Possibilities related to prediction by correctness")

Reassuringly it appears the mannequin was, on common, much less assured about its classifications for the wrong outcomes than the proper ones. (Though, the pattern dimension is simply too small to say something definitively.)

Let’s see what actions the mannequin had the toughest time with utilizing a confusion matrix.

predictionPerformance %>% 
  group_by(fact, predicted) %>% 
  summarise(depend = n()) %>% 
  mutate(good = fact == predicted) %>% 
  ggplot(aes(x = fact,  y = predicted)) +
  geom_point(aes(dimension = depend, coloration = good)) +
  geom_text(aes(label = depend), 
            hjust = 0, vjust = 0, 
            nudge_x = 0.1, nudge_y = 0.1) + 
  guides(coloration = FALSE, dimension = FALSE) +
  theme_minimal()

We see that, because the preliminary visualization prompt, the mannequin had a little bit of hassle with distinguishing between LIE_TO_SIT and LIE_TO_STAND lessons, together with the SIT_TO_LIE and STAND_TO_LIE, which even have related visible profiles.

Future instructions

The obvious future path to take this evaluation could be to aim to make the mannequin extra normal by working with extra of the provided exercise sorts. One other fascinating path could be to not separate the recordings into distinct ‘observations’ however as an alternative preserve them as one streaming set of knowledge, very like an actual world deployment of a mannequin would work, and see how nicely a mannequin may classify streaming knowledge and detect modifications in exercise.