"Could not allocate arena" after adding Bluetooth functions to voice recognition code - tensorflow

Question

I am trying to make a Bluetooth speaker controlled by voice commands. I have successfully written code to connect to the speaker via Bluetooth. Then I have successfully trained a neural network and got the speaker to recognize my commands (Speaker has a microphone). I followed this tutorial and code: https://github.com/atomic14/voice-controlled-robot/tree/main/firmware.
However, when I try to combine the two codes together, I get a "could not allocate arena" error and the code doesn't run properly. All I did was add "#include <btAudio.h>", "btAudio audio = btAudio("ESP_Speaker");", and "audio.begin();" to the main file of the voice recognition code. I have attached code for the main file. All of the other files are pretty much unchanged. In addition, I've attached a picture of the output after the program uploads: Code output. Does anyone have an idea on how I should approach fixing this error? Thank you.
#include <Arduino.h>
#include <WiFi.h>
#include <driver/i2s.h>
#include <esp_task_wdt.h>
#include "I2SMicSampler.h"
#include "ADCSampler.h"
#include "config.h"
#include "CommandDetector.h"
#include "CommandProcessor.h"
#include <btAudio.h>
#include <stdint.h>
#include <esp_avrc_api.h>
btAudio audio = btAudio("ESP_Speaker");
// i2s config for using the internal ADC
i2s_config_t adcI2SConfig = {
.mode = (i2s_mode_t)(I2S_MODE_MASTER | I2S_MODE_RX | I2S_MODE_ADC_BUILT_IN),
.sample_rate = 16000,
.bits_per_sample = I2S_BITS_PER_SAMPLE_16BIT,
.channel_format = I2S_CHANNEL_FMT_ONLY_LEFT,
.communication_format = I2S_COMM_FORMAT_I2S_LSB,
.intr_alloc_flags = ESP_INTR_FLAG_LEVEL1,
.dma_buf_count = 4,
.dma_buf_len = 64,
.use_apll = false,
.tx_desc_auto_clear = false,
.fixed_mclk = 0};
// This task does all the heavy lifting for our application
void applicationTask(void *param)
{
CommandDetector *commandDetector = static_cast<CommandDetector *>(param);
const TickType_t xMaxBlockTime = pdMS_TO_TICKS(100);
while (true)
{
// wait for some audio samples to arrive
uint32_t ulNotificationValue = ulTaskNotifyTake(pdTRUE, xMaxBlockTime);
if (ulNotificationValue > 0)
{
commandDetector->run();
}
}
}
void setup()
{
Serial.begin(115200);
audio.begin();
delay(1000);
Serial.println("Starting up");
// make sure we don't get killed for our long running tasks
esp_task_wdt_init(10, false);
// start up the I2S input (from either an I2S microphone or Analogue microphone via the ADC)
#ifdef USE_I2S_MIC_INPUT
// Direct i2s input from INMP441 or the SPH0645
I2SSampler *i2s_sampler = new I2SMicSampler(i2s_mic_pins, false);
#else
// Use the internal ADC
I2SSampler *i2s_sampler = new ADCSampler(ADC_UNIT_1, ADC_MIC_CHANNEL);
#endif
// the command processor
CommandProcessor *command_processor = new CommandProcessor();
// create our application
CommandDetector *commandDetector = new CommandDetector(i2s_sampler, command_processor);
// set up the i2s sample writer task
TaskHandle_t applicationTaskHandle;
xTaskCreatePinnedToCore(applicationTask, "Command Detect", 8192, commandDetector, 1, &applicationTaskHandle, 0);
// start sampling from i2s device - use I2S_NUM_0 as that's the one that supports the internal ADC
#ifdef USE_I2S_MIC_INPUT
i2s_sampler->start(I2S_NUM_0, i2sMemsConfigBothChannels, applicationTaskHandle);
#else
i2s_sampler->start(I2S_NUM_0, adcI2SConfig, applicationTaskHandle);
#endif
}
void loop()
{
vTaskDelay(pdMS_TO_TICKS(1000));
}