minor edits

Author: ES-Renesas

examples/detectnet_v2/readme.md

@@ -70,17 +70,6 @@ The results of training can be seen in table below:
 | Retrained model  | ~31 MB     | 40%            | Retrained the pruned model with 100 epochs on KITTI dataset            |
 ------------------
 
-
-The AI inference on RZ/V boards is as follows:
-
-| Board   | Inference speed (ms) |
-|---------|-----------------------|
-| RZ/V2L  | ~ 220                |
-| RZ/V2H  | ~ 30                 |
-
-> **Note** : The following results are based on DRP-AI TVM v2.5. RZ/V2L also requires a patch to offload conv layer from CPU -> DRP-AI accelerator
-
-
 A snippet of inference on board via a pre-recorded slideshow is shown below:
 
 ![Detectnet V2 inference output](../../docs/assets/Detectnet_v2_result_1.png)

examples/segformer/readme.md

@@ -64,14 +64,7 @@ The results of training after 300 epochs can be seen in table below:
 | accuracy | 0.9       | 0.9    | 0.9      |
 
 
-AI inference on RZ/V boards for this model runs entirely on the CPU, as it is not currently supported by the DRP-AI accelerator. As a result, inference performance is signifcantly lower compared to models optimized for DRP-AI:
-
-
-| Board   | Inference speed (ms) |
-|---------|----------------------|
-| RZ/V2L  | ~ 9000               |
-| RZ/V2H  | ~ 3479               |
-
+AI inference on RZ/V boards for this model runs entirely on the CPU, as it is not currently supported by the DRP-AI accelerator. As a result, inference performance is lower compared to models optimized for DRP-AI.
 
 > **Note:** Currently an embedded bug has been found in the postprocessing that causes dogs to be consistenly infered as, the following issue will be fixed in the next release of the tool.