chore: Fixed improper table display

Some tables were not displaying as intended. The problem has been fixed as of this commit.

Author: Dewalade1

README.md

@@ -285,19 +285,19 @@ The scenarios used include:
 
 1. CPU Hardware with Model precision FP16
 
-    command used:
+ command used:
 
-    ```bash
-    python3 src/main.py --input_type demo --precision high --batch_size <batch size choice>
-    ```
+```bash
+python3 src/main.py --input_type demo --precision high --batch_size <batch size choice>
+```
 
 2. CPU Hardware with Model precision FP32
 
 command used:
 
-    ```bash
+```bash
     python3 src/main.py --input_type demo --face_model intel/face-detection-adas-0001/FP32/face-detection-adas-0001 --face_landmark_model intel/landmarks-regression-retail-0009/FP32/landmarks-regression-retail-0009 --head_pose_model intel/head-pose-estimation-adas-0001/FP32/head-pose-estimation-adas-0001 --gaze_model intel/gaze-estimation-adas-0002/FP32/gaze-estimation-adas-0002 --precision high --batch_size <batch size choice>
-    ```
+```
 
 The results of these benchmark scenarios can be found in the Results section
 
@@ -321,6 +321,7 @@ This sections contains results of benchmark tests on the model performed with th
 | CPU  | FP32 | 100 | 0.11 | 0.67 | 0.78 | 0.31  | 16.28 |
 | CPU  | FP16 | 100 | 0.07 | 0.59 | 0.66 | 0.34  | 13.47 |
 
+
 #### Summary of Results
 
 From the benchmarking results above, We can deduce that the models initialize and load faster at FP16 precision. This faster models load time shortens the application's initialization time allowing the application to start up faster at FP16 model precision than when the models are at FP32 precision. The inference speed falls within a range of values but it peaks at the batch size of 1 for both precisions. Also, The average inference speed is lowest when the batch size is 50. The total inference time of the application spikes when batch size is 1 and drops to about 1.30s at bach size 25 and slowly declines as the batch size is increased.  
@@ -437,21 +438,29 @@ Microarchitecture Usage: 23.2%
 | sched_yield | libc.so.6 | 0.140s |
 | [Others]	| N/A* | 0.928s |
 
+
+
 #### Top Tasks
 
 | Task Type | Task Time | Task Count |
+|-----------|-----------|------------|
 | void cv::utils::logging::LogTagManager::assign(const string&, cv::utils::logging::LogTag*)| 0.000s | 1 |
 | cv::utils::logging::LogTag* cv::utils::logging::LogTagManager::get(const string&) |	0.000s | 1 |
 
+
+
 #### Effective CPU Utilization Table
 
 | Simultaneously Utilized Logical CPUs | Elapsed Time | Utilization threshold |
+|--------------------------------------|--------------|-----------------------|
 | 0 | 5.3801632009 | Idle |
 | 1	| 2.280570021 | Poor |
 | 2 |	0	|  Ideal |
 
 ![Effective CPU Utilization](./img/effective-cpu-utilization.png)
 
+
+
 #### Summary of Results
 
 The VTune Amplifier was used to perform an analysis on the inference pipeline to fine the top hotspots in the application. The data collection and collation was performed in 7.661s, 2.180s out of which was used by the CPU to actively execute the application. A total of 21 threads were used by the CPU. The Hotspots insights section of the summary showed that the application utilized only 1 CPU core with a 13.3% parallelism and a microarchitecture usage of 23.2%.
@@ -482,7 +491,8 @@ The application also logs the Gaze vector outputs and inference statistics in th
 
 - Batch count: Number of batches processed before inference was finished or interrupted.
 
-#### Logged Last outputs
+
+#### Final outputs that were logged
 - Face coordinates: Coordinates of the first face in the frame detected by the face detection model.
 
 - Left eye coordinates: Coordinates of the left eye of the face detected by the facial landmark model.